Soria Preceedings

See
discussions, stats, and author profiles for this publication at: https://www.researchgate.net/publication/303864063
The role of nitrogen availability on methane

abatement biotechnologies
Conference Paper · June 2016
CITATIONS READS
0 237
3 authors:
Juan Carlos López Raul Muñoz

Universidad de Valladolid Universidad de Valladolid
11 PUBLICATIONS 52 CITATIONS 152 PUBLICATIONS 3,306 CITATIONS
SEE PROFILE SEE PROFILE
Guillermo Quijano
Universidad Nacional Autónoma de México
61 PUBLICATIONS 768 CITATIONS
SEE PROFILE
Some of the authors of this publication are also working on these related projects:
Generation of gaseous biofuels from wastes View project
Algae biofuels and biorefinery View project
Available from: Juan Carlos López

Retrieved on: 28 September 2016
Towards a Sustainable Development
Soria , Spain
June 6 & 7, 2016
Sponsors
Welcome
On behalf of the Scientific Programme and Organizing
Committees, we are delighted to welcome you to the 1st
International Conference on Bioenergy & Climate Change
– Towards a Sustainable Development. We also wish to
welcome you to Soria, one of the greenest regions in Spain
and Europe.
The conference will focus on the latest developments on

bioenergy production and available technologies against
climate change, bringing together students, professionals
from academia, industry and public administrations. We
hope that you will find opportunities to contribute,
exchange, learn and interact with world experts in this
exciting and important field.
Dr. Guillermo Quijano Dr. Alberto Ordaz
Organizing Committee Chairs
1
International Scientific Committee
Fernando Fdz-Polanco – University of Valladolid (Spain)
Umberto Desideri – Università degli Studi di Pisa (Italy)
Luc Malhautier – École des mines d’Alès (France)
Carmen Gabaldón – University of Valencia (Spain)
Fédéric Thalasso – Cinvestav-IPN (Mexico)
Israel Díaz – University of Valladolid (Spain)
Evelyne Touraud – École des mines d’Alès (France)
Francisco Omil – University of Santiago de Compostela (Spain)
Flor Cuervo – UAM-Iztapalapa (Mexico)
Luis Miguel Romeo – University of Zaragoza & CIRCE (Spain)
Raúl Muñoz – University of Valladolid (Spain)
Francesco Fantozzi – Università degli Studi di Perugia (Italy)
Madalena Alves – University of Minho (Portugal)
Luis Hernández – University of Valladolid (Spain)
Alfons Stams – Wageningen University (Netherlands)
Alberto Ordaz – Mexiquense University (Mexico)
Pilar Lisbona – University of Valladolid (Spain)
Jorge Gómez – UAM-Iztapalapa (Mexico)
Daphne Hermosilla – University of Valladolid (Spain)
Pietro Bartocci – Università degli Studi di Perugia (Italy)
Pedro García – University of Valladolid (Spain)
Theo S.O. Souza – University of Sao Paulo (Brazil)
Guillermo Quijano – University of Valladolid (Spain)
2
Organizing Committee
Susana Calvo Ruiz – University of Valladolid
Adolfo Mercado Santamaría – University of Valladolid
José Ángel Miguel Romera – University of Valladolid
Luis Bonilla Morte – University of Valladolid
Ivonne Figueroa González – University of Valladolid
Daphne Hermosilla Redondo – University of Valladolid
Pilar Lisbona Martín – University of Valladolid
Epifanio Díez Delso – University of Valladolid
Israel Díaz Villalobos – University of Valladolid
Susana Gómez Ortega – Scientific Park, University of Valladolid
Alberto Ordaz Cortés (Chair) – Universidad Mexiquense del Bicentenario
Guillermo Quijano Govantes (Chair) – University of Valladolid
3
Program at a glance
Sunday 5th
Time
Welcome cocktail (a registration desk will be available) – Universitary Residence “Duques de Soria”
20:00-21:30
Monday 6th
Time
Registration & Coffee
8:00-9:00
9:00-9:20 Opening Notes (Guillermo Quijano)
KEYNOTE by Alfons Stams - Wageningen University (Netherlands)
9:20-9:50
“Innovative anaerobic microbial processes to conserve energy and produce biofuels”
ORAL PRESENTATIONS – BIOENERGY PRODUCTION
9:50-10:10 A novelty approach in air pollution control: Recycling VOC emissions into bioenergy
Daniel Bravo, Josep Manuel Penya-Roja, F. Javier Álvarez-Hornos and Carmen Gabladón
Structural insights on Ni-Ce-Zr catalysts for hydrogen production in biogas model streams by direct
10:10-10:30 catalytic reforming
Manuel Benito, Lucía Rodríguez and Gonzalo Muñoz
Production of biodiesel starting from a continuous thermal reactor
10:30-10:50
Sarabia-Martinez D.E., Silva-Rodriguez A., Vargas-Mondragón L.Z., Larios-Ortíz E.A., Ruelas-Ramirez E.H.
Dissolved Methane Recovery from an Anaerobic Effluent Using a PDMS Hollow Fiber Membrane
10:50-11:10 Contactor
María Henares, Marta Izquierdo, Carmen Gabaldón and Vicente Martínez-Soria
11:10-11:40 Coffee break
ORAL PRESENTATIONS – RESIDUES VALORIZATION I
11:40-12:00 Assessing the potential of industrial waste streams for volatile fatty acids (VFA) production
Garcia-Aguirre Jon, Aymerich Enrique and Esteban-Gutiérrez Myriam
Crude glycerol use as carbon source in a sulfate-reducing UASB for sulfur recovery from S-rich effluents
12:00-12:20
Mabel Mora, Javier Lafuente, Jordi Fernández, Ricard March, David Gabriel
Fluidised bed gasification of sewage sludge using different bed materials
12:20-12:40 Virginia Pérez López, José María Murillo Laplaza, Raquel Ramos Casado, Alfonso Pascual Delgado, José
María Sánchez-Hervas
The role of nitrogen availability on methane abatement biotechnologies
12:40-13:00
Juan Carlos López, Raúl Muñoz, Guillermo Quijano
13:00-13:50 Poster session
13:50-15:00 Lunch break
KEYNOTE by Raúl Muñoz - University of Valladolid (Spain)
15:00-15:30 “A techno-economical evaluation of physical/chemical and biological technologies for biogas
upgrading”
ORAL PRESENTATIONS – BIOGAS
15:30-15:50 Assessment of Thermal Cracking Tar from Sewage Sludge Gasification
Isabel Ortiz González, Manuel Jesús Benito González and José María Sánchez Hervás
Feedback Control of H2S Concentration in an Anoxic Biotrickling Filter
15:50-16:10
Javier Brito, Fernando Almenglo, Martín Ramírez, José M. Gómez, Domingo Cantero
Application of MBR to convert H2 and CO2/biogas into bioCH4
16:10-16:30
Natalia Alfaro, Israel Díaz, Fernando Fdz-Polanco and María Fdz-Polanco
Application of feedback and feedforward control strategies in aerobic biotrickling filters for biogas
desulfurization
16:30-16:50
Luis R. López, Mabel Mora, Caroline Van der Heyden, Eveline Volcke, Juan A. Baeza, Javier Lafuente,
David Gabriel
16:50-17:00 Closing Notes (Alberto Ordaz)
Visit to the San Saturio hermitage & picnic
18:00
Meeting point: main entrance of the Agrarian Engineering School
4
Tuesday 7th
Time
Registration & Coffee
8:00-9:00
9:00-9:20 Opening Notes (Guillermo Quijano)
KEYNOTE by Umberto Desideri - Università degli Studi di Pisa (Italy)
9:20-9:50
“Overview of the renewable energy penetration in Italy and Europe – challenges and perspectives”
ORAL PRESENTATIONS – ALGAL BACTERIAL SYSTEMS
Kinetic Exploration Of The Microalgal Growth Rate Of Species Residing In A Waste Stabilization Pond
9:50-10:10
Bjorge Decostere, Andres M. Alvarado, Esteban A.M. Sanchez, Diederik Rousseau, Ingmar Nopens and
Stijn WH Van Hulle
Aquatic plants and microalgae biomass for the production of biofuels within a biorefinery
10:10-10:30 Eugenia J. Olguín, Gloria Sánchez-Galván, Ricardo E. González-Portela, Daniel A. García-López and Erika
Rodríguez-Barradas
Effect of pH, temperature and P/N ratio over a microalgae consortium for biomass, lipids and chlorophyll
10:30-10:50 production
Yaremi López Hernández, Refugio Rodríguez Vázquez, Jorge I. Chairez Oria, Luis G. Torres Bustillos
Production of Chlorella vulgaris in culture media with different source of nitrogen and carbon to obtain
10:50-11:10 lipids
Citlally Ramírez-López, Jorge Isaac Chairez-Oria, Luis Carlos Fernández-Linares
11:10-11:40 Coffee break
Masterclass by Theo S.O. Souza - University of São Paulo (Brazil)
11:40-12:10
“Mathematical Modeling as a Powerful Tool for the Optimization of Biogas Production”
ORAL PRESENTATIONS – RESIDUES VALORIZATION II
Availability of agro residues and estimation of the theoretical, sustainable and technical energetic
12:10-12:30
potentials in Mexico: The scenario in the last ten years
Jessica Denisse Valle García, Danahé Marmolejo Correa and Edgar Vázquez Núñez
The activities on thermal degration of biomass at the Biomass Researche Centre, Perugia, Italy
12:30-12:50
Pietro Bartocci, Gianni Bidini, Paolo Laranci, Michele D'Amico, Francesco Fantozzi
Biomass blending as a way to reduce nox emissions during the combustion of biomass residues
12:50-13:10
Brice Piednoir, Jean-Michel Commandré, Anthony Benoist, Sarah El Fassi, Gilles Vaitilingom
Copper bioleaching from waste printed circuit boards by Acidithiobacillus ferrooxidans
13:10-13:30
Eva Benzal, Montserrat Solé, Concepció Lao-Luque, Xavier Gamisans, Antonio David Dorado
13:30-14:00 Poster session
14:00-15:00 Lunch break
KEYNOTE by Luc Malhautier - École des mines d’Alès (France)
15:00-15:30 “New insights from the microbial ecology approach in anaerobic bioreactors and hydrogen-
fermenting systems”
ORAL PRESENTATIONS – STRATEGIES AGAINST CLIMATE CHANGE
15:30-15:50 Fluidized Bed Oxy-Combustion of Coal and Biomass: an Expedient Way to Reduce Pollutant Emissions
Carlos Lupiáñez, Luis Ignacio Díez
Lanthanum modified hydrotalcites for CO2 capture in sorption enhanced water gas shift (sewgs)
15:50-16:10 processes
Yarima Torreiro, Marta Maroño, Juan García, José María Sánchez
Life Cycle Assessment of biogas desulfurization by different technologies to valorize the Biogas potential
16:10-16:30 energy in sewage plants
Patricio I. Cano, Joan Colón, Martín Ramírez, Javier Lafuente, David Gabriel and Domingo Cantero
Nitrous oxide: global trends, mitigation and abatement
16:30-16:50
Osvaldo D. Frutos, Raquel Lebrero, Guillermo Quijano, David Marín, Raúl Muñoz
16:50-17:10 Closing Notes (Alberto Ordaz)
Tapas night at the main square of Soria
20:00
Meeting point: main entrance of the city’s park “Alameda de Cervantes”
5
Monday June 6th 2016 – Keynote speakers
9:20-9:50 Alfons Stams. Innovative anaerobic microbial processes to conserve energy and
produce biofuels
Speaker: Alfons Stams studies the anaerobic microbiology of environmental biotechnological

processes such as waste and wastewater treatment, soil and groundwater remediation,
production of fuels and biochemical and recovery of metals. Major contributions were made to
understand the functioning of syntrophic communities of anaerobic bacteria and methanogenic
archaea, sulfate- and sulfur-reducing bacteria and bacteria that respire with organic and
inorganic chlorine compounds. Alfons Stams published about 400 peer reviewed articles and
book chapters on the physiology and application of anaerobic microorganisms. In 2012 he
obtained an advanced ERC grant to study novel anaerobes for a biobased economy. A main
research strategy is to rapidly explore the physiological potential of newly isolated
microorganisms by genomics and proteomics approaches. In 2013 he received in a consortium
with 4 other Dutch scientists the prestigious Gravity Grant of the Netherlands Ministry of
Education, Culture and Science to study Anaerobes for Health and Environment. This research is
embedded in the newly founded Soehngen Institute for Anaerobic Microorganisms
(www.anaerobic-microbiology.eu).
Synopsis: Unlike aerobic microbial processes which result in the degradation of organic
compounds to CO2, anaerobic microbial processes allow to conserve energy and produce
organic compounds. Currently, much attention is paid to produce organic chemicals and biofuels
from low-grade organic waste materials. In the presentation two topics will be addressed. One
topic is energy conservation and product formation from lipids waste materials. Anaerobic
conservation of lipids waste materials is much less studied than anaerobic conservation from
lignocellulosic waste materials. It will be discussed how after hydrolysis long chain fatty acids
can be efficiently converted to form biogas. Glycerol, the other constituent of lipids, can be
fermented by anaerobic microorganisms to a set of interesting compounds, such as ethanol, 1,3-
propanediol and succinate. The properties of some novel glycerol-fermenting bacteria and their
potential for biotechnological application will be discussed. The other topic is microbial product
formation from syngas. By gasification any kind of naturally-occurring and synthetic organic
material (e.g. lignocellulose and plastic and rubber) can be converted to syngas, which is an
energy-rich gas mixture composed of mainly carbon monoxide, carbon dioxide and hydrogen.
This syngas can be used by anaerobic microorganisms that are able to form organic chemicals
and biofuels. It will be discussed how synthetic microbial communities of anaerobes can be used
to form biogas and high value chain elongated products. The research approach is to combine
bacteria with good syngas-converting properties with bacteria that form interesting products.
Key references
Cavaleiro AJ, Pereira MA, Guedes AP, Stams AJM, Alves MM, Sousa DZ (2016) Conversion of Cn-uinsaturated into Cn-2-saturated LCFA
can occur uncoupled from methanogenesis in anaerobic bioreactors. Environmental Science & Technology 50:3082-3090.
Diender M, Stams AJM, Sousa DZ (2015) Pathways and bioenergetics of anaerobic carbon monoxide fermentation. Frontiers in
Microbiology 6, article 1275.
Spirito CM, Richter H, Rabaey K, Stams AJM, Angenent LT (2014) Chain elongation in anaerobic reactor microbiomes to recover
resources from waste. Current Opinion in Biotechnology 27:115-122.
Stams AJM, Plugge CM (2009) Electron transfer in syntrophic communities of anaerobic bacteria and archaea. Nature Reviews
Microbiology 7:568-577
Van Gelder AH, Sousa DZ, Stams AJM, Sánchez-Andrea I (2014) Ercella succinigenes gen. nov., sp. nov., a new anaerobic succinate-
producing bacterium. International Journal of Systematic and Evolutionary Microbiology 64:2449-2454
6
15:00-15:30 Raúl Muñoz. A techno-economical evaluation of physical/chemical and
biological technologies for biogas upgrading
Speaker: Raúl Muñoz obtained his degree in Chemical Engineering at the University of Valladolid
(Spain) in 2001 and completed his Doctoral Thesis “Algal-bacterial photobioreactors for the
degradation of toxic organic pollutants” at the Dept. of Biotechnology of Lund University. In
2005, Raúl joined the Department of Chemical Engineering and Environmental Technology at
University of Valladolid, where he initiated the research lines on biological gas treatment and
algal-bacterial processes. Raúl has 15 years of experience in the development of innovative algal-
bacterial processes for wastewater treatment. His research in the area of biological gas
treatment has focused on the optimization of two-phase partitioning bioreactors for industrial
VOC treatment, biological odour treatment, development of high-mass transfer bioprocesses
for GHG treatment, and on the upgrading and bioconversion of biogas. The applicant has 117
works published in JCR scientific journals, 70% of them within the Q1 of the JCR classification,
and has co-authored 10 International Book Chapters. His works have been cited >2500 times,
with an H index of 28 (Scopus ID: 16319160700). He has supervised 10 PhD Thesis (+7 on-going)
and 16 Master Thesis. Finally, he has been Principal Investigator in 19 R&D projects.
Synopsis: The absence of fiscal incentives for the use of biogas in most countries entails the
need for an optimization in the upgrading of biogas in order to promote the utilization of this
renewable energy source. The presence of contaminants in biogas such as CO2, H2S, H2O, N2, O2,
methyl siloxanes and halocarbons requires nowadays the implementation of costly upgrading
processes. Conventional biogas upgrading is based on physical/chemical technologies providing
CH4 purities of 88-98% and removal efficiencies for H2S, halocarbons and methyl siloxane
removals higher than 99%. Unfortunately, their high energy and chemical requirements limit the
environmental and economic sustainability of physical/chemical biogas upgrading technologies.
In this context, biological processes have emerged in the past decade as an economic and
environmentally friendly alternative to conventional biogas upgrading technologies. Thus,
biotechnologies such as microalgae-based CO2 fixation, H2-assisted litoautotrophic CO2
bioconversion to CH4, enzymatic CO2 dissolution or fermentative CO2 reduction can provide
consistent CO2 removals of 80-100% with CH4 purities of 88-100%, while allowing the valorization
of CO2 into bioproducts of commercial interest. Similarly, H2S removals over 99% are consistently
achieved in aerobic and anoxic biotrickling filters, algal-bacterial photobioreactors and digesters
under microaerobic conditions. In addition, recent investigations have shown the potential
biodegradability of methyl siloxanes and halocarbons under both aerobic and anaerobic
conditions. This keynote will provide a state of the art comparison of physical/chemical and
biological technologies for the removal of CO2, H2S, halocarbons and methylsiloxanes from
biogas.
Key references
Lebrero R, Toledo-Cervantes A.L, Muñoz R, del Nery V, Foresti E (2016) Biogas upgrading from vinasse digesters: a comparison
between an anoxic biotrickling filter and an algal-bacterial photobioreactor. Journal of Chemical Technology &
Biotechnology, DOI: 10.1002/jctb.4843.
Muñoz R, Meier L., Diaz I, Jeison D (2015) A critical review on the state-of-the-art of physical/chemical and biological technologies for
an integral biogas upgrading. Reviews in Environmental Science and Bio/Technology 14:727-759.
Posadas E, Serejo M, Blanco S, Perez R, Garcia-Encina P, Muñoz R (2015) Minimization of Biomethane Oxygen Concentration during
Biogas Upgrading in Algal-Bacterial Photobioreactors. Algal Research 12: 221-229.
Serejo M, Posadas E, Boncz M, Blanco S, Garcia-Encina P, Muñoz R (2015) Influence of biogas flow rate on biomass composition
during the optimization of biogas upgrading in microalgal-bacterial processes. Environmental Science & Technology
49(5): 3228-3236.
Bahr M, Diaz I, Dominguez A, Sanchez-Gonzalez A, Muñoz R (2014) Microalgal-biotechnology as a platform for an integral biogas
upgrading and nutrient removal from anaerobic effluents. Environmental Science & Technology 48(1): 573-581.
7
Tuesday June 7th 2016 – Keynote speakers
9:20-9:50 Umberto Desideri. Overview of the renewable energy penetration in Italy and Europe
– challenges and perspectives
Speaker: Umberto Desideri hods MSc and PhD degrees in Mechanical Engineering and Energy
Engineering from the University of Florence, respectively, being currently a full professor of
Thermal Machines at the University of Pisa, Italy. Professor Desideri is member of the Editorial
Board of the ASME Journal of Fuel Cell Science and Technology and Associate Editor of Applied
Energy, Applied Thermal Engineering and Energy Conversion and Management. He is founder of
the Spin-off company “Tecnologie per la riduzione delle emissioni engineering srl”. His research
focuses on innovative and high efficiency fossil fired power generation systems, renewable
energy systems (e.g. geothermal, concentrated solar power, PV and biomass), hydrogen and
fuel cells, carbon capture/storage, energy saving in buildings and industry, tri-generation and
polygeneration, energy efficient buildings, building integrated renewable energy systems, life
cycle assessment and carbon footprint.
Synopsis: The presentation will start with an overview of the renewable energy penetration in
Italy and Europe, highlighting the challenges that this scenario has created for grid stability and
energy price volatility. In electric grids with significant penetration of intermittent renewables, it
is becoming a critical issue to balance the supply and demand at sunrise and even more at
sunset or when significant weather changes modifiy the wind production in short times. Some
current trends for future development and the potential technologies to cope with the
problems that we are now facing will be listed and discussed. The role of biomass among the
other renewable energy sources will be discussed and some specific proposals for the
development of bioenergy and biofuels will be addressed. In this regard, bioenergy can play a
role for a secure and efficient energy mix in the future by considering the potential to use it in
energy storage, heat and power plants, mobility and grid balancing. Most current technologies
could find new interest if we look at them from different points of view than it is currently done.
Key references
Cinti, G., Baldinelli, A., Di Michele, A., Desideri, U. Integration of Solid Oxide Electrolyzer and Fischer-Tropsch: A sustainable pathway
for synthetic fuel (2016) Applied Energy, 162, pp. 308-320.
Barelli, L., Desideri, U., Ottaviano, A. Challenges in load balance due to renewable energy sources penetration: The possible role of
energy storage technologies relative to the Italian case (2015) Energy, 93, pp. 393-405.
Cinti, G., Desideri, U. SOFC fuelled with reformed urea (2015) Applied Energy, 154, pp. 242-253.
Antonelli, M., Desideri, U. The doping effect of Italian feed-in tariffs on the PV market (2014) Energy Policy, 67, pp. 583-594.
Cinti, G., Desideri, U., Penchini, D., Discepoli, G. Experimental analysis of SOFC fuelled by ammonia (2014) Fuel Cells, 14 (2), pp. 221-
230.
8
11:40-12:10 Theo S.O. Souza. Mathematical Modeling as a Powerful Tool for the Optimization of
Biogas Production
Speaker: Dr. Theo is an assistant Professor at the Department of Hydraulic and Environmental
Engineering, Polytechnic School, University of São Paulo (Brazil). PhD in Hydraulic Engineering
and Sanitation (São Carlos Engineering School, University of São Paulo, 2011), with research
topics on nitrogen removal processes applied to anaerobically pretreated domestic sewage,
mainly autotrophic denitrification using sulfide as electron donor. Postdoctoral position at the
University of Valladolid, Spain (2012-2013), working with the implementation of the ADM1 to
evaluate thermal hydrolysis pretreatment of sludge. Postdoctoral position at the University of
São Paulo, Brazil (2014), focused on landfill leachate treatment. Expertise in biological processes
for wastewater treatment, anaerobic digestion modeling, alternative biotechnologies for
nitrogen removal and carbon-nitrogen-sulfur interactions in biosystems. Current research areas
include nitrogen removal from landfill leachate coupling nitritation and anammox in single-stage
reactors, and ADM1 applications for modeling the anaerobic digestion of organic wastes.
Synopsis: Anaerobic digestion has been extensively used for treating a wide range of substrates,
including several types of organic solid wastes, industrial and domestic sewage. Methane-rich
biogas production consists of a major advantage to justify the focus on development and
optimization of systems based on anaerobic biotechnology, mainly when alternative sources of
energy are an impending necessity. In this context, anaerobic digestion has evolved from
traditional applications such as mixed-digesters treating WWTP sludge to a new set of reactor
configurations treating a new set of wastes, including low-strength wastewater and co-
digestion of different substrates. Mathematical modeling has been employed as a tool to assess
the complex environment provided by anaerobic digestion and its resulting biogas production,
and has also historically evolved from simple to large structured models, culminating in the
creations of IWA's Anaerobic Digestion Model No 1 (ADM1), which was continuously studied and
improved during the last 15 years. New anaerobic configurations and substrates require updates
for anaerobic models to cover bioprocesses that were not considered before in anaerobic
systems, novel hydrodynamics aspects and substrate characteristics. This presentation focuses
on an up-to-date overview of such topics, including traditional and novel potentials for
anaerobic systems, a brief history of anaerobic modeling and general aspects of the ADM1,
recent applications and modifications of anaerobic models, as well as challenges, constraints
and future perspectives on the modeling and optimization of biogas production.
Key references
Batstone DJ, Virdis B (2014) The role of anaerobic digestion in the emerging energy economy. Current Opinion in Biotechnology
27:142-149.
Batstone DJ, Puyol D, Flores-Alsina X, Rodríguez J (2015) Mathematical modelling of anaerobic digestion processes: applications and
future needs. Reviews in Environmental Science & Biotechnology 14: 595-613.
Donoso-Bravo A, Mailier J, Martin C, Rodríguez J, Aceves-Lara CA, Wouwer AV (2011) Model selection, identification and validation in
anaerobic digestion: a review. Water Research 45:5347-5364.
Lauwers J, Appels L, Thmpson, IP, Degrève J, Van Impe JF, Dewil R (2013) Mathematical modelling of anaerobic digestion of biomass
and waste: Power and limitations. Progress in Energy and Combustion Science 39:383-402.
Souza TSO, Carvajal A, Donoso-Bravo A, Peña M, Fdz-Polanco F (2013) ADM1 calibration using BMP tests for modeling the effect of
autohydrolysis pretreatment on the performance of continuous sludge digesters. Water Research 47:3244-3254.
9
15:00-15:30 Luc Malhautier. New insights from the microbial ecology approach in anaerobic
bioreactors and hydrogen-fermenting systems
Speaker: Luc Malhautier is a senior researcher at the “Ecole des mines d'Alès” (French institute
for engineering and applied science). His research interests fall under the overall area of
biologically-based processes for environmental protection. His main research area focuses on
biological treatment of gaseous emissions and the application of an integrated approach
coupling functional analysis (removal efficiency of gas pollutants) and microbial community
characterization (DGGE, qPCR). The long-term evaluation of the robustness of air treatment
technologies exposed to transient disturbances (shock loads) has been investigated by the
research group of professor Malhautier, including the determination of resistance and resilience
capacities and the demonstration of the memory effect developed by microbial communities.
Bioprocess intensification has been also examined for the complete removal of gas pollutants
such as H2S, which is a common contaminant of biogas. His research activity has generated
many key publications on his research field. Most recently, in the bioenergy context, research on
the valorization of volatile compounds has been done by his research group in Alès.
Synopsis: Even though they are the key actors of bioprocesses, little is known about the
microbial communities involved in such bioreactors: how are they influenced by the
environmental conditions, and how do they drive the ecosystem function? Unravelling the
biotic/abiotic interactions is a challenging issue that can provide control, diagnostic and
prevision tools for process monitoring. The biotic component in anaerobic bioreactors and
hydrogen-fermenting systems has been more and more investigated during the last decades,
thanks to the rapid development of molecular tools. Even though not exempted from potential
bias, they provide a more representative image of the community, without restricting to
cultivable cells. Current techniques for the characterization of microbial communities in such
engineered ecosystems will then be briefly presented. Despite the large amounts of data
generated by using molecular biological tools and genomic, transcriptomic, and proteomic
approaches, gaining mechanistic insights and using the data for predictive and practical
purposes is still a huge challenge. Nevertheless, some results concerning the biodiversity-
ecosystem function relationship in the context of bioenergy are relevant and will be highlighted
in the second part of the presentation. We will pay special attention to the start-up period of
bioreactors, spatial and temporal dynamics of microbial community structure and function, and
shifts of microbial community structure in response to perturbation (transient change).
Key references
Cabrol L, Malhautier L (2011) Integrating microbial ecology in bioprocess understanding: the case of gas biofiltration. Applied
Microbiology & Biotechnology 90:837-849.
de los Reyes FL III, Weaver JE, Wang L (2015) A methodological framework for linking bioreactor function to microbial communities
and environmental conditions. Current Opinion in Biotechnology 33:112-118.
Rittmann BE, Krajmalnik-Brown R, Halden RU (2008) Pre-genomic, genomic and postgenomic study of microbial communities
involved in bioenergy. Nature Reviews Microbiology 6:604-612.
Tolvanen KES, Karp MT (2011) Molecular methods for characterizing mixed microbial communities in hydrogen-fermenting systems.
International Journal of Hydrogen Energy 36:5280-5288.
Vanwonterghem I, Jensen PD, Ho DP, Batstone DJ, Tyson GW (2014) Linking microbial community structure, interactions and
function in anaerobic digesters using new molecular techniques. Current Opinion in Biotechnology 27:55-64.
10
1st International Conference on Bioenergy & Climate Change - Towards a Sustainable Development
Oral Presentations – Bioenergy Production

Monday June 6th 2016
11
Soria, Spain – June 6 & 7, 2016
A novelty approach in air pollution control:

Recycling VOC emissions into bioenergy
Daniel Bravo1&2, Josep Manuel Penya-Roja1, F. Javier Álvarez-Hornos1 and Carmen
Gabladón1
1Reasearch Group GI2AM, Department of Chemical Engineering, University of Valencia, Avd.
Universitat s/n, 46100, Spain. Email: Daniel.bravo@uv.es
2Pure Air Solutions b.v, P.O. Box 135, 8440 AC Heerenveen, The Netherlands
ABSTRACT
This study aims to evaluate the anaerobic bioscrubber, an environmentally friendly solution to
control air waste emissions at facilities while decreasing its energy demand through the
production of bioenergy. For this purpose, a 15-m3 prototype was operated over two years.
The plant couples a water scrubber and a high-rate mesophilic anaerobic reactor. In this
system, volatile organic compounds (VOCs), coming from solvent use, are transferred from
waste gases to the washing liquid in the scrubber, allowing to accomplish the strict European
regulation. The scrubbing liquid containing the dissolved solvents is fed to the reactor for its
biological degradation by anaerobic granular sludge. This step permits to reuse the washing
liquid in the scrubber, and simultaneously produces biogas that can partially replace the
consumption of energy generated by fossil fuel at the facility. The results obtained in this study
proved the robustness of this technology, the scrubber unit achieved removal efficiency higher
than 90%, with an average inlet VOC concentration of 1348 ± 427 mg-C Nm-3 (63 ± 3%wt of
ethanol, 24 ± 5%wt of ethyl acetate and 9 ± 5%wt of 1-ethoxy-2-propanol). The performance of
the reactor with a 95 ± 3% removal efficiency was stable spite of the shift organic load during
nights and weekends. The in-situ upgraded biogas resulted in a methane content of 95 ± 2%vol.
This is the first study at industrial scale showing the great potential of anaerobic bioscrubber
to control VOC air emissions by recycling them into renewable energy (biogas).
Keywords: Anaerobic reactor, Bioenergy, Bioscrubber; VOC, air pollution control.
1. INTRODUCTION
The control of volatile organic compounds from air emissions has gained importance
in European Union environmental policy over the past decades. The flexographic sector, which
represents the 17% of the European printing technologies, contributed around 0,3% of the
total turnover in 2003 (Ernst&Young and Intergraf, 2007),and it is one of the industries affected
by the environmental legislation on VOC air emissions into atmosphere since most inks
contain solvents (VOC). The VOCs emission of the these facilities are characterized by large
volumes of off gases that contain low concentrations of oxygenated compounds (Datta and
Philip, 2012), thus processes like incineration become cost intensive due to the requirement
of addition fuel for higher efficiency, being the bioprocess the best suit technologies to control
VOC emissions (Kennes et al., 2009).
Among biological techniques, a new alternative is anaerobic bioscrubber, where VOCs
are transferred from waste gases to the washing liquid in a scrubber unit and the scrubbing
liquid containing the dissolved solvents is fed to an expanded granular sludge bed (EGSB)
reactor for its biological degradation. This step continuously regenerates the washing liquid
prior to being recycled into the scrubber. Despite its complexity, main advantages are, in
comparison with aerobic biodegradation system, lower operation cost, smaller space
requirement and methane production, which can be used as a renewable energy source for
12
industrial use (Mudliar et al., 2010; Van Lier, 2008). Although the two units are well-studied
technologies for VOC emissions control (Ozturk and Yilmaz, 2006) and packaging wastewater
treatment (Lafita et al., 2015), there is no data available for the application of the both
technologies coupled to treat VOCs emissions.
The proper VOCs abatement in this technology relies on the satisfactory coupling of both main
units, thus the transfer of the VOCs in the scrubber while the microbial population and physical
integrity of the anaerobic sludge granules developed properly thanks to the maintenance of
the pH and nutrients concentration in optimum range, promoting at the same time microbial
activity for the degradation of the solvents.
The present work provides the first example of successful long-term operation of an industrial
scale anaerobic reactor controlling the emissions of VOCs originated in a flexographic facility
(Altacel Transparant Verpakkingsind, The Netherlands)
2. MATERIALS AND METHODS

The University of Valencia and Pure Air Solutions b.v. (Netherlands) developed an anaerobic
bioscrubber industrial plant (BONCUSTM, number of patent: ES254257R1). In Figure 1, the
conceptual diagram of the industrial unit can be found.
Biogas
Clean air Purge
CENTRIFUGAL
PUMP ANAEROBIC
REACTOR
CENTRIGUAL
FAN
PUMP
Polluted air SCRUBBER
Centrifugal pump
Fresh water Nutrients Chemical pH control
Figure 1: Conceptual diagram of anaerobic bioscrubber pilot system.
The industrial plant setup was provided with a programmable logic controller (PLC) with
Twinsoft software (Servelec technologies, United Kingdom) to monitor and control the system.
Air and liquid flowrates, biogas flowrate, water and air temperature, pH, conductivity and water
levels were continuously monitored. The scrubber was retrofitted with a flame ionization
detector (FID) type gas analyser (model RS 53-T, Ratfisch Analysensysteme, Germany) to
continuously monitor the VOC concentration.
The scrubber was operated in countercurrent mode, VOC polluted air coming from the factory
was introduced below the column by mean of a blower and the water used to treat the air
stream was pumped by a centrifugal pump and collected in the scrubber tank, from where it
is pumped to the anaerobic reactor (8.7 m3) for solvent degradation. The anaerobic reactor
was supplemented with nutrients and pH was chemically kept under control. From day 0 to
day 484, the trial was divided into two stages in order to test two different packing material in
the scrubber column; the packing material A was installed in the stage 1 and the packing
13
material B was used in the stage 2. The experimental conditions of each stage are shown in
Table 1.
Table 1. Experimental plan for the removal of VOCs in a bioscrubber.
Stage-1 Stage-2
Days 0 - 95 96 - 130
Packing material A B
Specific surface, m-1 150 125
Water/air volume ratios 0.0035 - 0.0091 0.0076 - 0.0101
3. RESULTS AND DISCUSSION
3.1 Characterization of the foci

The site, where the anaerobic bioscrubber was located, operates on two shift productions (16
h) per day, 6 days per week, therefore the VOC feeding regime of the system was subject to
these working hours, since during the night and weekend closure periods there was no VOC
emission. The inlet emissions are quite stable, average daily values with standard deviation
of 1348 ± 427 mg-C·Nm-3 for the whole experimental period, the total emission during one day
changed depending upon the orders being printed. Regarding the composition, ethanol (60-
65% in mass), ethyl acetate (20-25% in mass) and ethoxy propanol (10-15% in mass) were
the main compounds; ¡Error! No se encuentra el origen de la referencia. presents all the
compounds in the air emission along with SD.
3.2 Plant Monitoring

In order to know the scrubber performance, a study based on the removal efficiencies with
different ratios of water and air flows were carried out for each packing material. The Figure 2
reported the efficiency achieved for each packing material, the broken line indicates the day
14
when the packing material B was installed and vertical lines show changes in the air water
flows ratio.
Figure 2. Performance of the scrubber with two packing material and working as a spray column.
The results show the expected tendency in the removal efficiency, which increased with higher
values of liquid to air flow rate. For packing A, the removal efficiency ranged between 78% for
the lowest water and air flow ratio applied (0.0035) and 97% for the highest ratio (0.0091). In
the case of the B packing material, the lowest removal efficiency achieved was 77% for the
ratio 0.00101 and the highest was 89% for ratio 0.0076. Regarding the emission with A
packing material trial, it was mainly composed by ethyl acetate (Henry constant of 6.5 M atm -
1
) and high soluble compounds such as ethanol (Henry constant of 200 M atm -1) were fully
removed for all tested conditions. Regarding the operational conditions, the pressure drop in
both cases was kept in feasible values.
The water with the solvents was pumped to the EGSB reactor for solvent degradation prior to
recirculate to the scrubber. The Figure 3 shows the weekly average of the organic loads and
the removal efficiency and in Figure 4 is depicted the weekly average of the volatile fatty acids
(VFA), the broken line indicate the switching from A to B packing material in the scrubber in
both Figures.
15
Figure 3. Weekly average of organic load and removal efficiency of the anaerobic reactor for the
testing period ( ͦ removal efficiency and ●organic load).
Figure 4. Weekly average of Volatile fatty acids in the effluent of the anaerobic reactor for the whole
experimental period.
The removal efficiency of the anaerobic reactor was calculated by mass balance taking into
account the amount of solvent transferred to the system in the scrubber and the removed
solvents with the daily purge. The removal efficiency was over 90% for the testing period,
showing the feasibility of the proposed technology for the VOC control. VFA were kept below
250 mg acetic acid ·L-1, except from day 55 to day 70 when high organic loads were applied,
revealing that the maximum load that could be treated in the anaerobic reactor is 3.5 kg COD
h-1. The chemical pH system controlled the pH in average of 7.4. Biogas was obtained with a
methane concentration of 95 ± 2%vol, which permits use it directly on-site without further
upgrading
4. CONCLUSIONS
The performance for VOC abatement of an industrial-scale anaerobic bioscrubber was studied
on-site at a flexographic facility. The technology was able to treat the emissions, transferring
the VOC in the scrubber unit and treating the dissolved solvents in the EGSB. The satisfactory
coupling of both units allows to keep the parameters of the water in optimum range for the
production of biogas with high content of methane. This work proves the robustness and
capability of the anaerobic bioscrubber for VOC removal from industrial air emissions,
obtaining at the same time bioenergy that can partially replace the consumption of energy
generated by fossil fuel at the facility.
ACKNOWLEDGMENTS
The research leading to these results has received funding from the People Programme (Marie
Curie Actions-ITN) of the European Union’s Seventh Framework Programme FP7/2007-2013/
under REA grant agreement n°606942 (TrainonSEC). Authors gratefully acknowledge the
financial support by Generalitat Valenciana (Spain, Prometeo project 2013/53). The authors
would like to give special thanks to Wim Schaap (Heineken, The Netherlands) for providing
the granular sludge.
16
REFERENCES
Datta, A. and Philip, L. (2012). Biodegradation of volatile organic compounds from paint
industries. Applied Biochemistry and Biotechnology, 167(3), 564–580.
Ernst&Young and Intergraf. Competitiveness of the European Graphic Industry. European
Communities, Brussels, Belgium (2007).
Gabaldón, C., Peñarrocha, J.M., Alvarez-Hornos, F.J. and Waalkens, A. (2015).
Procedimiento para la depuracion de gases que contienen compuestos orgánicos
volátiles. Patent No: ES2542257R1. Oficina Española de patentes y Marcas (España)
Kennes, C., Rene, E. R. and Veiga, M. C. (2009). Bioprocesses for air pollution control.
Journal of Chemical Technology and Biotechnology, 84(10), 1419–1436.
Lafita, C., Penya-roja, J. M. and Gabaldón, C. (2015). Anaerobic removal of 1-methoxy-2-
propanol under ambient temperature in an EGSB reactor. Bioprocess and Biosystems
Engineering, 2137–2146.
Mudliar, S., Giri, B., Padoley, K., Satpute, D., Dixit, R., Bhatt, P., Pandey, R., Juwarkar, A. and
Vaidya, A. (2010). Bioreactors for treatment of VOCs and odours - a review. Journal of
Environmental Management, 91(5), 1039–54.
Ozturk, B. and Yilmaz, D. (2006). Absorptive Removal of Volatile Organic Compounds from
Flue Gas Streams. Process Safety and Environmental Protection, 84(5), 391–398.
Van Lier, J. B. (2008). High-rate anaerobic wastewater treatment: Diversifying from end-of-
the-pipe treatment to resource-oriented conversion techniques. Water Science and
Technology, 57(8), 1137–1148.
17
Structural Insights on Ni-Ce-Zr Catalysts for

Hydrogen Production in Biogas Model Streams by
Direct Catalytic Reforming
Manuel Benito1, Lucia Rodríguez2 and Gonzalo Muñoz2

1CIEMAT, Unit for Energy Valorization of Fuels and Wastes, Av. Complutense 40, 28040 Madrid,
Spain. mj.benito@ciemat.es
2CSIC, Institute of Catalysis and Petrochemistry, C/ Marie Curie, 2. Cantoblanco, 28049 Madrid,
Spain, paralucia2@yahoo.es, gonzalo.munnoz@gmail.com
ABSTRACT
Biogas represents the highest potential among renewable energy sources. From this point of
view, the energy valorization of biogas streams into hydrogen could be an important way to
obtain renewable hydrogen from a waste, overcoming the negative environmental effects of
biogas emissions to the atmosphere. For this propose, a series of Ni-Ce-Zr catalysts have
been prepared modifying nickel concentration, and Ce/Zr ratio. Changes in catalysts structure
were identified by XRD. The catalysts with a 70 wt%Ce show a cubic crystalline phase while
the catalysts with a 70 %of Zr show a tetragonal structure, which have an important influence
on catalyst performance. This facts have an important effect on lattice parameter and therefore
on oxygen mobility which is related with coking resistance. Reaction tests were performed in
differential conditions and CH4/CO2 1:1 on catalyst samples with different composition.
Catalysts characterization by XRD, SEM, N2-Adsorption isotherms and ICP was performed on
fresh samples. Post reaction samples were characterized by temperature programmed
oxidation and scanning electron microscopy in order to identify and quantify the possible coke
formed. The effects of structural modifications on catalysts performance, this means methane
conversion, H2/CO, CH4/CO2 and finally coke deposition on the catalysts, are the focus of this
work.
Keywords: Biogas; Catalyst; Reforming; Hydrogen.
1. INTRODUCTION
Hydrogen production from hydrocarbons seems to be the best option to achieve a progressive
transition to the ‘hydrogen economy’ in the near- and mid-term scenario. Nowadays, a 48% of
the World’s hydrogen production is obtaining by steam reforming, being natural gas the major
feedstock. However, there are alternative processes also used for hydrogen production that
in some cases can be more advantageous. From this point of view, with the increasing interest
in renewable energy sources and the reduction of greenhouse gases emission, CO2-methane
reforming has attracted a great deal of attention as an alternative to the use of biogas as fuel
[1]. Nowadays, it is important to note that biogas represents the highest potential among
renewable energy sources. Taking advantage of biogas generated in sewage treatment
plants, farms, landfill sites, organic solid waste digesters, organic liquid effluent treatment
plants, and methanization plants not only it avoids emission of gases with high global warming
18
potential but also it produces energy as a fuel for heat and power equipment. However, the
use of biogas for electricity generation by internal combustion engines has some drawbacks
such as the production of certain contaminants directly related to the combustion process
(NOx, SOx, dioxins, furans, particulates) and low electrical efficiency, even lower for small
power systems. For the reforming of enriched CO2 biogas streams, a process that is
particularly interesting is ‘dry’ reforming with carbon dioxide. Dry methane reforming allows
the use of the remainder CO2 in the biogas mixture as an oxidant in the reformer without
expensive methane separation processes [2]. The ‘Dry reforming’ reaction is an endothermic
reaction, and requires operating temperatures near 900 ºC to achieve high conversion that
usually provokes catalyst deactivation by coke formation and sintering of metallic phase and
support [3,4]. There are two commercial technologies based on this reaction: Calcor process
of Caloric GmbH [5] and the Sparg process (Sulphur passivated Reforming) of Haldor Topsøe
[6]. Dry reforming process is often combined with steam reforming or partial oxidation to
operate under more favorable conditions and to reduce carbon deposition on catalysts, which
represents one of the most important problems in this process.
With regard to catalyst development, the active phases for CO2 methane reforming are metals
that belong to VIII Group with the exception of Os. Commercial nickel based catalysts are
usually supported on alumina or magnesium aluminate and contain nickel in proportions
between 10 and 20 wt. %, and to a lesser extent other components that give stability against
sintering and coke deposition in the reaction conditions. However, in the design of catalysts
for reforming, is important not only the choice of the metallic phase or promoter, but also the
support. The latter can play, depending on its chemical and structural properties, a role not
only in catalytic activity, but also in catalyst stability against coke deposition, since the metal-
support synergistic action can alter the mechanism of the reaction. The accumulation of coke
deposited produced by methane decomposition could be controlled by gasification rate, and
therefore on the availability of surface oxygen species [7]. From this point of view, the results
obtained with supports based on cerium and zirconium oxides are promising, showing higher
OSC than ceria with higher thermal stability and redox ageing [8,9].
The aim of the present work is to evaluate the influence of the catalysts structure in Ni-CeO2–
ZrO2 catalysts on their surface and catalytic properties in the reaction of CH4 reforming with
CO2 for hydrogen production, that simulates the relative CH4/CO2 composition presented in a
biogas stream without any diluting gas, studies that are not usually reported in literature [10].
X-ray diffraction (XRD), temperature-programmed reduction (TPR), thermo-gravimetric
analysis (TGA) and Scanning Electron Microscopy (SEM) were the technical methods used
for the samples characterization.
2.1. Catalysts preparation
Six catalysts (two series) based on Ni-Ce-Zr with different nickel content and Ce/Zr ratio, were
prepared by one step sol-gel method. Zirconium iso-butoxide (Zr(OC4H9)4, 80% in 1-butanol),
Cerium nitrate (Ce(NO3)3•6H2O), Nickel nitrate (Ni(NO3)2•6H2O) and 1-butanol were the
precursor salts and solvent used to prepare the samples, in concentration necessary to obtain
a 5 wt. % of Ni and Cerium composition of 20 wt. %, 60 wt. % and 80 wt. %, respectively. The
pseudo-gel obtained was submitted to drying for 1h at 60 º C. After 60 h the catalyst precursor
was heated up to 120 ºC at heating rate of 3 ºC·min-1. After that, the catalyst precursor was
19
calcinated at 750 ºC. Prior to catalyst testing, catalysts were reduced in a 50 mLN·min-1
hydrogen stream for 2 h at 700 ºC.
2.2. Catalysts characterization
Chemical composition of the catalysts was determined by plasma emission technique (ICP-
AES). XRD patterns of the prepared catalysts were obtained a diffractometer X´PERT PRO
PANALYTICAL. Measurements were performed at 2ϴ range of 2-100º, using CuKαβ radiation
(λ=1.540598 Å) and removing Kαβ radiation by a nickel filter, with a 0.05º·s-1 scanning and an
accumulation time of 2s. Crystalline phases were identified by means of X'PERT HIGHSCORE
PLUS software, comparing the patterns obtained with Joint Committee on Powder Diffraction
Standards 1998 data base.
Nitrogen adsorption-desorption isotherms at -196 ºC were obtained automatically by means

of a Micromeritics ASAP 2100 equipment. Powder samples were previously out-gassed at 150
ºC for 12 h. Specific surface area was calculated from isotherms by mathematical regression
using Brunauer-Emmet-Teller equation.
Temperature programmed reduction (TPR) tests were performed in a METTLER-TOLEDO

TGA/SDTA 851 thermo-balance, by introducing a catalyst sample (c.a. 50 mg) in a quartz pan.
The experiment was carried out with 10% H2-90% N2 (50 mLN·min-1) stream as the reactive
gas and 50 mLN·min-1 of N2 as the purge gas, in the temperature range of 25-900 ºC at heating
rate of 10 ºC·min-1.
Temperature programmed oxidation (TPO) analyses of catalysts after reaction, were carried
out in a METTLER-TOLEDO TGA/SDTA 851 thermo-balance, in order to determine the
amount of carbon deposited on catalysts. The standard protocol involved the mass change of
samples (50 mg) in a quartz pan during its heating in a 50 mLN·min-1 of N2 stream used as
purge gas and a 20% O2-80% N2 (50 mLN·min-1) stream used as reactive gas from 25 to 900
ºC at 10 ºC·min-1.
The images of scanning electron microscopy (SEM) were obtained in a scanning electron
microscope Hitachi S-3000N. Chemical analyses were performed with an analyzer by energy
dispersive X-ray of Oxford Instruments, model-sight INCAx.
2.3. Activity tests
Biogas simulated reforming reaction experiments were carried out in a MICROACTIVITY

REFERENCE PID ENG&TECH equipment with a fixed-bed tubular quartz flow reactor (4 mm
i.d., 330 mm length) operating isothermally at 700 ºC and at atmospheric pressure. Catalyst
samples (20-100 mg, grain size of 0.42-0.50 mm) were mixed with inert diluter (SiC) in 1:4
weight ratio. In every run, the catalysts were flushed by nitrogen flow at 700 ºC, and
subsequently, CH4-CO2 mixture in 1:1 ratio was fed into the reactor at a flow rate of 100
mLN·min-1. The reaction products were analyzed with an AGILENT 6890N chromatograph
connected in line, equipped with TCD and FID detectors. Helium was used as carrier gas.
Methane and carbon dioxide conversion, denoted as XCH4 and XCO2 and the effluent gas
composition, denoted as Ci, were calculated from chromatograph data.
20
3.1 Structural Characterization
The XRD patterns of the 5 wt. % Ni series synthesized catalysts are compiled in Fig. 1. The
Ce/Zr ratio used in one step sol-gel process has led to structural changes in the catalysts
synthesized. The addition of Ce shifts the peaks to lower 2Θ values, producing an increase of
interplanar spacing which is expected as Zr4+ ions are replaced by the bigger Ce4+ [11,12].
The catalysts with 20 wt.% and 80 wt.% of cerium oxide present an XRD pattern that
corresponds to Ce-Zr mixed oxides with composition very similar to the nominal composition
of the catalysts (Ce0.12Zr0.88O2, JCPDS 82-1398, Ce0.75Zr0.25O2, JCPDS 28-0271) that
correspond to tetragonal phase and face-centered cubic structure respectively, although it is
almost impossible to discriminate the existence of both phases in zirconia or cerium enriched
mixed oxides based only in XRD [13]. In the case of the sample of composition 60 wt.% Ce
(molar ratio Ce:Zr 1:1) a segregation of phases was observed, obtaining an enriched cerium
phase and another zirconium oxide-enriched phase. Vegard's law is widely used to confirm
the presence of solid solutions. Differences in ionic radius of Ce4+ (0.97 nm) and Zr4+ (0.84
nm) provide a linear relationship between lattice parameters and Ce content in solid solution
for the cubic samples. In the tetragonal samples, this linear relationship is established with c
lattice parameter. Figure 2 represents c lattice parameter calculated by indexing structural
planes compiled from JCPDS to XRD sample patterns vs. Ce content obtained by ICP-AES
analysis. Despite different structures (cubic and tetragonal) being considered, high linearity
(r2=0.93) is maintained between the experimental points representing the samples of
synthesized catalysts and the c value represented for reference samples of CeO2 (JCPDS 34-
0394) and ZrO2 (JCPDS 24-1164), with cubic and tetragonal structure, respectively. The
samples with 20 wt.% cerium, show a greater difference between the c parameter determined
by XRD and the value estimated by the Vegard´s law. This fact could indicate a partial
substitution of Ce4+ by Ni2+ in the lattice [14].
3.2. Catalytic Activity
Reaction tests were performed at 700 ºC and CH4:CO2 1:1, conditions where coke deposition
is favored thermodynamically. Catalysts were operated at WHSV at 300,000 L•kg -1•h-1 much
higher than values reported in literature (6,000-190,000 L•kg-1•h-1), to force the catalyst to
operate at low conversion and to evaluate its performance in severe operation conditions to
analyze possible deactivation phenomena.
21
Figure 3 shows the activity calculated as mmoles of methane or carbon dioxide converted by
nickel mass after 25 h in reaction conditions. In all samples studied, carbon dioxide conversion
was higher than methane conversion as equilibrium estimation predicted due to reverse water
gas shift contribution. It is a remarkable fact that activity of the catalysts with higher nickel
content was lower that catalysts with 5 wt. % of Ni following the same trend in both catalysts
series. In order to discuss the results and to establish correlations with catalyst deactivation
and post-reaction characterization, XCH4/XCO2, H2/CO, H2/CH4 and H2O/CO2 ratios from
product distribution were calculated (Fig. 4). The presence of water in products distribution
(figure not shown) is a consequence of reverse water gas shift reaction (1) contribution,
favored thermodynamically at high temperatures (700 ºC).
CO2+H2CO+H2O ΔHº=41.1 kJ•mol-1 (1)
The catalysts with 5 wt. % of Ni showed higher XCH4/XCO2 ratio vs. catalysts with higher
nickel content. By comparison of values obtained with those calculated using thermodynamic
Gibbs model (XCH4/XCO2=0.55), the experimentally obtained values for catalysts with 5 wt.%
of Ni exceed the theoretically obtained values by minimizing the Gibbs free energy. Moreover,
the H2/CO obtained was also higher than predicted by equilibrium estimations (0.42). The
results obtained could be explained on the basis of considering the contribution of methane
thermal decomposition (2).
CH4  2H2+ C (s) ΔHº=74.82 kJ•mol-1 (2)
Nevertheless, the catalysts with 5 wt. % of Ni presented a good stability with low deactivation
rate (results doesn´t showed), especially the samples 5Ni60CeZr and 5Ni80CeZr, where the
methane conversion decrease rates were 0.024 and 0.020 % per hour respectively. The
increase of cerium content not only maintains the catalyst activity but also decrease the
deactivation rate significantly. Catalyst series with 10 wt. %Ni presented lower XCH4/XCO2
ratio. However, this catalyst series presented methane conversion higher than 5 wt. % Ni
series which suggests an increase in CO2 conversion ratio. Possible processes that
corroborate this hypothesis are the contribution of reverse water gas shift reaction (1), and the
reverse Bouduard´s reaction (3). An increase of H2/CO ratio when Ce content increased was
22
also observed. Water formation as a consequence of reverse water gas shift reaction was
slightly higher for 10% Ni series.
C (s)+ CO22CO ΔHº=172.2 kJ•mol-1 (3)
3.3. Post-reaction characterization
Figure 5 shows the mass loss and DTG curves obtained for 5 wt. % Ni series in a TPO
experiment. By means of DTG deconvolution, two contributions to coke gasification were
identified that were ascribed to coke deposits α-type and β-type formed on Ni particles and
support respectively. Figure 6 compiles qualitative and quantitative information extracted from
TPO analysis performed on catalysts samples (5%Ni series and 10%Ni series) after 25 h in
reaction conditions. Coke formation rates for every sample were calculated dividing the total
coke gasificated in TPO experiments by time in reaction conditions of the catalytic samples. It
is important to note that no graphitic coke was detected in TPO experiments that indicates that
coke deposits could be removed easily in a catalyst regeneration process.
The 5 wt. % Ni series presented much lower coke deposits formation rate that were identified
as α-type, coke deposited on Ni particles and easily removed in a low temperature oxidation
process. Catalyst 5Ni20CeZr was the sample that presented the lowest amount of coke
deposits and therefore the lowest coke formation rate. Thus, the increase of Ni content favors
coke formation that is promoted by Ce concentration, increasing the β-type coke (amorphous)
formation rate.
2.0
100.5 0.00001 Carbon 
Carbon 
100.0
0.00000
1.5
99.5
Coke formation rate /%/h
-0.00001
99.0
Mass /%
DTG
1.0
98.5 -0.00002
98.0
-0.00003
0.5
97.5
5Ni20CeZr
5Ni60CeZr -0.00004
97.0
5Ni80CeZr
0.0 r
0 100 200 300 400 500 600 700 800 900 1000
Zr eZ
r r eZ Zr Zr
Temperature / ºC Ce C C eZ 0C Ce Ce
0 60 2 0 0
i2 i i 80 Ni i6 i8
5N 5N 5N 10 10
N
10
N
Figure 5. TG-DTG curves for 5wt.% sample Figure 6. Amount of coke determined for 5
series wt.% series by TG experiments
4. CONCLUSIONS
The one-step sol-gel process used to prepare Ni-Ce-Zr catalysts had a great influence on
catalyst structure and a determining role on catalyst performance. Tetragonal structure in
catalyst prepared with cerium content of 20 wt.% was obtained. The increase of cerium content
up to 80% produced a structural modification, obtaining a fluorite type structure. Catalysts with
cerium content of 60 wt. % resulted in formation of a segregated structure. Differences in
reducibility, Ni integration and interaction in Ce/Zr structure were detected as a function of
Ce/Zr ratio that lead to different reaction performance in the presence of a simulated biogas
23
stream with a CH4:CO2 ratio 1:1 without inert gas feeding. Catalyst series with 5 wt.% Ni
presented the best activities regarding nickel content, being 5Ni80CeZr the catalyst with
lowest deactivation rate. Post reaction TPO experiments put in evidence the negligible carbon
formation on 5%Ni series after 25 h in reaction conditions.
REFERENCES
[1] The state of renewable energies in Europe. 10th EurObserv’ER Report, December 2010, ISSN 2101-
9622.
[2] Lau C S, Tsolakis A, Wyszynski M L (2011), Biogas upgrade to syngas (H2-CO) via dry and oxidative
reforming, Int J Hydrogen Energy 36: 397-404.
[3] Rostrup-Nielsen J R, Bak Hansen J-H (1993 ), CO2-reforming of methane over Transition Metals, J
Catal: 144:38.
[4] Wang S, Lu G Q (1998), Reforming of methane with carbon dioxide over Ni/Al2O3 catalysts: Effect
of nickel precursor, Appl Catal 169: 271-280.
[5] Teuner S C, Newman P, CO through CO2 reforming: The Calcor Standard and Calcor Economy
Processes, F Von Linde, Oil Gas European magazine 3: 44-46.
[6] Rostrup-Nielsen J R (2002), Syngas in perspective, Catal Today: 71 (3-4): 243-247.
[7] Hu Y H, Ruckenstein E (2004) , Catalytic Conversion of Methane to Synthesis Gas by Partial
Oxidation and CO2 Reforming, Adv Catal; 48: 297-345.
[8] Yao H C, Yu Yao Y F (1984), Ceria in automotive exhaust catalysts: I. Oxygen storage, J Catal 86:
254-265.
[9] Roh H S, Potdar H S, Jun K W, Kim J W, Oh Y S (2004), Carbon dioxide reforming of methane over
Ni incorporated into Ce–ZrO2 catalysts, Appl Catal A 276: 231.
[10] Corthals S, Van Nederkassel J, Geboers J, Winne H, Van Noyen J, Mohens B, Sels B, Jacobs P
(2008), Influence of composition of MgAl2O4 supported NiCeO2ZrO2 catalysts on coke formation
and catalyst stability for dry reforming of methane, Catal Today 138: 28-32.
[11] Annegi E, de Leinterburg C, Dolcetti G, Trovarelli A (2006), Promotional effect of rare earths and
transition metals in the combustion of diesel soot over CeO 2 and CeO2–ZrO2, Catal Today 114:
40-47.
[12] Kuznetsova T G, Sadykov V A, Specific features of the defect structure of metastable nanodisperse
ceria, zirconia, and related materials (2008), Kinet Catal 49: 840-858.
[13] Damyanova S, Pawelec B, Arishtirova K, Martinez Huerta M V, Fierro J L G, The effect of CeO 2 on
the surface and catalytic properties of Pt/CeO2-ZrO2 catalysts for methane dry reforming (2009),
Appl Catal B: Environ 89: 149-159.
[14] Montoya J A, Romero-Pascual E, Gimon C, del Angel P, Monzón A (2000), Methane reforming with
CO2 over Ni/ZrO2-CeO2 catalysts prepared by sol-gel, Catal Today 63: 71-85.
24
Production of biodiesel starting from a

continuous thermal reactor
Sarabia-Martinez D.E., Silva-Rodriguez A., Vargas-Mondragon L.Z., Larios-Ortíz E.A.,
Ruelas-Ramirez E.H.
Universidad Mexiquense del Bicentenario, Unidad de Estudios Superiores Tultitlan, Av.
Ex Hacienda de portales s/n Col. Villa esmeralda, Tultitlán, Estado de México, C.P.
54910, México.
*Corresponding author: eh.ruelas@umb.mx, Tel. (044) 5520784737.
ABSTRACT
Biodiesel is a liquid fuel produced from vegetable oils or animal fats this is the best alternative
to replace partial or completely to the petroleum-derived diesel fuel, given the technical and
environmental advantages that offers. Biodiesel is obtained by a Transesterification that
depending on the process and catalyst may require up to 300 ° C for no longer than one hour
or close to 50 ° C for up to 12 hours, so the energy expenditure is quite high. On the other
hand, the production of biodiesel currently is done by batches, which raises the costs of
investment in infrastructure for its production. Against this background, it is essential to find
alternatives for obtaining biodiesel that will improve the energy balance and efficiency of the
process. A viable alternative to improve the rate of return energy in the production of biodiesel
is the solar thermal energy given that our country receives from the Sun in a single day, the
amount of energy that would be required for an entire year. In this work was the design of a
continuous reactor CSP for the Transesterification process in the production of biodiesel in
order to improve the rate of return energy in its production using solar thermal energy as well
as biodiesel continuously, and not by lot. The design of the reactor provides heating for the
transesterification reaction, being the maximum air working temperature 90 °C.
Keywords: biodiesel, energy, solar thermal, pressure and rate of energy return.
1. INTRODUCTION
Biofuels are currently in constant growth due to environmental issues and high prices for
hydrocarbons, so that different countries are developing new and innovative ways to get
different types of bio-fuels as well as renewable energy sources, which include bioethanol,
biogas, hydrogen and biodiesel [1]. Today there are different ways of producing biofuels,
notably biodiesel, although the biofuels industry is still a long way to go.
Conventional techniques for the production of biodiesel use reactors, as in the majority of
industrial chemical processes, where takes place a Transesterification reaction, which consists
of the transformation of one type of ester in another. When the original ester reacts with an
alcohol, the Transesterification reaction is called Alcoholysis [2]. Conventional techniques for
the production of biodiesel reactors are used [3], in either continuous or discontinuous (batch)
[4]. Although the efficiency of these devices is high, it has a limitation the long residence times,
high operating cost, efficiency of low production and high power consumption [5].
However to consider that the biodiesel will become a real energy alternative, is essential to
this present not only features equivalent to the retrieved from petroleum diesel, but also in the
whole of processes for obtaining, get positive energy and economic balances. A viable
alternative is solar thermal energy or solar heating, given that our country receives from the
Sun in a single day the amount of energy that would require all year round. Collecting the
adequate solar radiation, we can obtain heat and electricity. Among the advantages more
relevant and attractive solar thermal systems (STS), is that for its performance, is it just
depends on the incidence of solar radiation as energy resource, they do not require automatic
control systems in operation, which does not represent a problem for use in rural or remote
25
areas [6]. Can be used various technologies known as collectors or heliostats (large mirrors
with solar tracking systems), to focus and gather the sunlight and turn it into heat medium-high
temperature [7]. Some studies done in Paraguay, mentioned the use of thermal solar energy
(EST), from a cylindrical parabolic concentrator, for dehydration of foods. [8]. In Mexico the
national laboratory systems concentration and chemical Solar (LACYQS), is carrying out a
project that is divided into three sub - major projects: solar high temperature furnace, plant of
solar detoxification for water treatment and a field of heliostats tests [9]. The parabolic
cylindrical technology is a mature commercial technology, with an experience of over 20 years
of operation [10]. However, today there is information regarding the use of this type of
technology as a source of heat in chemical reactions, and much less for the obtaining of a
biofuel. For this reason the majority of the investigations are aimed at generating new
approaches in the process of production of biofuels, which do not represent greater damage
to the environment and which in turn generate an added value, becoming one of their biggest
challenges [3]. This proposal was intended, the design of a continuous reactor solar thermal
to obtain biodiesel. In addition, built reactor is inexpensive, so would conservation policy, with
energy saving to replace expensive equipment in the practice run, thus reducing both the cost
of installation and maintenance.
2. METODOLOGY
To develop this project 3 consistent fundamental stages were held in:

1) Obtaining biodiesel was performed by a method published by García Martínez (2008), part
of castor oil, with the purpose of having experimental results with which to compare the
performance of the designed system.
2) Construction of the system of solar concentration for catalysis of reactions consists in a
concentrator cylindrical Luminaire (CCP), capable of generating temperatures from 250 to 300
° C. This stage takes as a basis the methodology proposed by Luis Escobar Vidríales in 2007
[11] with measures.
3) For the design of the reactor continued CSP, were considered equations of a Tubular flow
Piston.se Reactor design considered heat transfer efficiency, was provided by means of a fluid
carrier thermos (motor oil), in the same way the following parameters were taken into account
for your design:
a) Volume, it is essential to know the volume of the interior of our reactor, to determine
the dimensions of our conductive tubes, where the reaction will take place, as well as
the external volume.
b) Residence time, is the ratio of the volume residing on the inner tubes and the time that
would soon take over the reactor, with an estimated time of 4 minutes
c) Specific heat of mixing
d) The flow of heat that we must maintain in the reaction, since it will be reflected in the
quality of biodiesel.
Later the reactor was built, so that inside has several glass tubes, where the Transesterification
reaction, is carried out and the outer pipe is steel material where the fluid termotransportador,
in conjunction with a mixture of the methoxide, methanol and catalyst feed tank. Suitable for
Design materials were selected with respect to what was reported by Cunningham [12].
3. RESULTS
In relation to the obtaining of the biodiesel from the method proposed by García Martinez, was
obtained 60% of biodiesel in performance compared to what was reported by bibliography. As
for the construction of the system of solar concentration for catalysis of reactions, was
determined that the ideal height is 15 cm, it is the point at which higher temperature is reached
from the surface of the reflective plate. The results of micro-reactor design are reflected based
on each of the parameters that are shown below. The volume that have internal reactor tubes,
consisting of a length of 1m and a diameter of 0.5 cm is:
26
𝑉 = 19.635 𝑐𝑚3
And equals the volume of the outer tube with respect to the fluid termo-transporterer, with a
length of 1 m and a diameter of 2 in a:
𝑉 = 2026.8346 𝑐𝑚3
Based on the foregoing only worked with the required volume counting occupied space that
was:
𝑉 = 1520.126 𝑐𝑚3
The flow which will have the reactor will be taken with a time of 4 min:
1520.126 𝑐𝑚3
𝑄=
4 𝑚𝑖𝑛
3
𝑄 = 380.0315 𝑐𝑚 ⁄𝑚𝑖𝑛
The number of tubes that works reactor, depended on the area of each of the inner
tubes as well as tube outer left a total of 13 internal glass tubes and thus make efficient
Transesterification reaction.
Fig. 1. Finished continuous reactor design.
The flow of heat from the reactor will depend on the specific heat of mixing which is:
𝐶𝑝𝑚𝑒𝑧 = 𝐶𝑝𝑡𝑟𝑖𝑔𝑙𝑖𝑐𝑒𝑟𝑖𝑑𝑜 + 𝐶𝑝𝑁𝑎𝑂𝐻 + 𝐶𝑝𝐶𝐻3 𝑂𝐻
𝐽 𝐽 𝐽
𝐶𝑝𝑚𝑒𝑧 = 200 ⁄𝐾𝑔°𝐾 + 1465.38 ⁄𝐾𝑔°𝐾 ∗ 0.04 + 2533 ⁄𝐾𝑔°𝐾 ∗ 3
𝐽
𝐶𝑝𝑚𝑒𝑧 = 758.5152 ⁄𝐾𝑔°𝐾
Calculation for the mass;
𝑚 = (∑ 𝜌 ∗ 𝑛𝑐𝑜𝑚𝑝𝑜𝑛𝑒𝑛𝑡𝑒𝑠 ) ∗ 𝑄
𝑔 𝑔
𝑚 = (5 𝑚𝑙⁄𝑚𝑖𝑛)(5.4254 ⁄𝑚𝑙 ) = 27.127 ⁄𝑚𝑖𝑛
27
Heat obtained:
𝑔 𝐽 𝐽
𝑞 = (27.127 ⁄𝑚𝑖𝑛) (7.8585 ⁄𝑔 °𝐾 ) (35 °𝐾) = 7461.22796 ⁄𝑚𝑖𝑛
The loss of the fear-carrier fluid temperature:

𝑞𝑡𝑟𝑖𝑔𝑙𝑖𝑐𝑒𝑟𝑖𝑑𝑜 = 𝑞𝑓𝑙𝑢𝑖𝑑𝑜 𝑡𝑒𝑟𝑚𝑜−𝑡𝑟𝑎𝑛𝑠𝑝𝑜𝑟𝑡𝑎𝑑𝑜𝑟
𝐽
7461.2279 𝐽 = (135017.91 𝑘𝑔)(1880 ⁄𝐾𝑔 °𝐾 )(346.13°𝐾 − 𝑇2 )
7461.2279 𝐽
𝑇2 = ( − 346.13°𝐾)
(135017.91 𝑘𝑔)(1880 𝐽⁄𝐾𝑔 °𝐾)
𝑇2 = 345 °𝐾
Temperature of the fluid termo-transporterer loss is 1 C, reliable result for what will be radical
temperature losses and an optimum temperature for reaction may be carried.
4. CONCLUSIONS
It can be concluded that in the process of Transesterification, that took place with the method
already reported in literature, obtained an efficiency of 60% of biodiesel. The parabolic
cylindrical collector (MCP) has the ability to capture and focus the thermal energy required for
the production of biodiesel. With respect to the continuous reactor thermal design, established
a relationship between the residence time of 4 min and a length of 1 m, according to the design
calculations determined that the CSP continuous reactor, has a capacity of 1.5 L and a flow
rate of 380.0315 cm3. As material of construction of the reactor, was chosen stainless steel
with a diameter of 2 in to the shell and glass for the inner tubes with an inside diameter of 0.5
cm. The energy required for heating of the reactor to carry out the reaction is 7.8585 J/g °K,
the heat lost is 1°C. So far, we can determine that the continuous reactor thermal reaches
temperatures enough to apply them in the process of Transesterification biodiesel production.
REFERENCES
[1] Secretaría de Energía , «Programa de Introducción de bioenergéticos,» SENER, México

D.F., 2012.
[2] 1. N. A. C. V. J. J. R. Iliana Ernestina Medina Ramírez, «biodiesel, un combustible
renovable,» redelyc, pp. 55,62-70, 2012.
[3] Y. Fernandez, O. Floiran y E. Maurelys, «Construcción de un reactor discontinuo para
la obtención de biodiesel a partir del aceite de Ricinus communis,» Avances en Ciencias
e Ingeniería, pp. 51-61, 2014.
[4] O. A. E. O. R. Duarte Vera, «Estudio de factibilidad de la producción de biodiesel en
reactores tanque agitado continuo de laboratorio,» Hidrogeno y Fuentes Sustentables
de Energía, 2011.
[5] L. a. L. W. Z. Qui, «Process intesification tecnologies in continuos biodiesel
producction,» Chemical Engineering and Processing Intensification, pp. 323-330, 2010.
[6] J. M. -. I. Ferrer, «Dimensionado de un sistema térmico solar mediante simulación y su
validacion energetica.,» Redalyc, pp. 55-65, 2013.
[7] Greenpeace, SolarPACES y ESTELA, «Energía Solar Térmica de Concentración
Perspectiva Mundial 2009,» Greenpeace, Madrid, 2009.
28
[8] C. Paredes Velasco , «Diseño de captador solar cilíndrico parabólico para aplicaciones
rurales en Paraguay,» Septiembre 2012.
[9] M. A. Duran Archuleta y J. L. Taddei Bringas, «Una perspectiva internacional sobre las
experiencias de manufactura de heliostatos para su desarrollo en México,» Energías
Renobables, pp. 2-6, 2013.
[10] Abengoa Solar, «Abengoa Solar,» 14 Marzo 2014. [En línea]. Available:
http://www.abengoasolar.com.
[11] L. G. Vidriales Escobar, «Colector de canal parabólico para generación directa de vapor
para calor de proceso,» 2007.
[12] R. E. y. L. J. L. CUNNINGHAM, Fundamentos del Diseño De Reactores, Buenos Aires:
Edeba, 1972.
29
Dissolved Methane Recovery from an Anaerobic

Effluent Using a PDMS Hollow Fiber Membrane
Contactor
María Henares*, Marta Izquierdo, Carmen Gabaldón and Vicente Martínez-Soria
Research Group GI2AM, Department of Chemical Engineering, University of Valencia, Avda.

Universitat s/n, 46100 Burjassot, Spain. Tel. +34 963543085, Fax: +34 963544898. www.uv.es/giam
*Corresponding author: maria.henares@uv.es
ABSTRACT
The objective of this work was to study the feasibility of a new technology for the recovery of
methane from the effluent of anaerobic reactors, to avoid greenhouse gas diffuse emissions
and enhance energy production. For this purpose, a polydimethylsiloxane (PDMS) hollow fiber
degassing membrane contactor was used to successfully remove/recuperate the dissolved
methane from the recirculating stream of a lab-scale Expanded Granular Sludge Bed
anaerobic reactor that treats a wastewater with ethanol at 25ºC. The performance of this
membrane contactor was evaluated at vacuum and sweep gas operations under different
conditions of liquid flow rate (QL, 0.36-10.80 L h-1), vacuum pressure (Pvac, 140-800 mbar), N2
flow rate (QN2, 2.7-27.0 L h-1), and mode of operation (lumen and shell side). At vacuum
conditions a raise of the CH4 recovery was achieved when QL and Pvac were decreased and
increased respectively, achieving maximum recovery efficiency (RE) of almost 80%. At sweep
gas conditions a negligible improvement in the RE was observed with the increase of QN2, and
a maximum RE of around 70 % was obtained. Experiments with the liquid passing through
lumen side showed higher RE than that of shell side. Results have been discussed from mass
transfer analysis, and experimental mass transfer coefficients have been compared with
different correlations. Results show degassing membrane contactors as a new promising and
viable technology to recover the dissolved methane from anaerobic effluents, with a positive
energy balance and a considerable reduction of carbon equivalent emissions.
Keywords: anaerobic bioreactor; biogas recovery; degassing membrane contactor;

greenhouse gas emissions.
1. INTRODUCTION
Anaerobic wastewater processes are a well-established and proven technology for the
treatment of various categories of industrial wastewaters (Cakir and Strestrom 2005). One of
the main advantages of anaerobic treatments, compared to aerobic or other systems, is the
production of methane, which can be used as fuel for generation of electricity or heat for
domestic and industrial use. Although most anaerobic wastewater treatments have been
usually conducted within mesophilic (30-40oC) or thermophilic (45-60oC) temperature ranges
(Lettinga et al., 2001), since anaerobic biodegradation is favoured at that conditions, new
trends are focused in treatments at psychrophilic conditions (<30oC) as much simpler systems
are needed and there is no need of heating the reactor. Nevertheless, low wastewater
temperature processes involve a significant quantity of residual dissolved methane (D-CH4)
present in the water effluent of the reactor, as methane solubility raises with the decrease of
temperature. The removal/recovery of this residual D-CH4 can be useful because it would 1)
represent a recover of potential energy, 2) reduce the risk of explosive atmospheres
generations (maintaining methane concentration in air below the lower explosive limit < 5%
30
v/v), and 3) decrease the very negative carbon footprint to the environment due to fugitive
emissions of CH4 (El-Fadel and Massoud, 2001).
In this context, degassing membrane (DM) contactors appear as an emergent technology that
is being used to remove dissolved gases in several industrial processes. The main advantage
of this technology is related with the fact that a gas and a liquid phase come in contact in the
pore of the membrane, without the need of dispersion of one phase into another, and therefore
avoiding typical problems of foaming, emulsions or flooding of conventional technologies. DM
contactors present other advantages over conventional dispersed phase contactors, such as
availability at high and low flow rates as they are modular, easiness of scale up and wide
range of capacities adding or removing membrane modules, high interfacial area per volume
unit and high efficiency.
Hollow fiber module is the most common configuration used in DM contactors in order to
remove gases such as CO2 and O2 from a liquid phase, so a considerable amount of studies
related with the removal or recovery of these gases can be found (Ito et al 1998, Leiknes 2001,
Peng et al 2008). Unfortunately, studies about the removal of D-CH4 from anaerobic effluents
are very scarce yet (Cookney et al. 2016, Bandara et al 2013), and further research in this
field is needed to improve and deepen in the knowledge and performance of this technology.
The objective of this work was to study the performance of a commercial polydimethylsiloxane
(PDMS) hollow fiber degassing membrane module for the recovery of dissolved methane from
the effluent of an anaerobic reactor.

A commercial hollow fiber membrane contactor module (PDMSXA-250, supplied by
PermSelect®, MedArray Inc., USA) was selected as usual in water degassing industrial
applications. Main properties of DM module are summarized in Table 1.
Table 2. Characteristics of non-porous polydimethylsiloxane (PDMS) contactor module

Number of fibers 320
Effective length, m 0.083
Inner diameter, μm 190
Outer diameter, μm 300
Internal Area (Ai), m2 0.0159
External Area (Ae), m2 0.0250
Shell tube inner diameter, m 0.016
Packing fraction 0.113
A laboratory scale EGSB anaerobic reactor was operated at 25 ºC during more than 24
months. The EGSB reactor was initially inoculated with 4 L of granular anaerobic sludge from
the wastewater treatment plant of a local brewery. The reactor treated 8 L d-1 of a synthetic
wastewater polluted with ethanol with an organic load rate of 32 kg chemical oxygen demand
(COD) m-3d-1. A high recirculation flow was kept to expand the sludge bed with an up-flow
velocity of 10.7 m h-1. A liquid-gas separator device was placed at the top of the reactor, and
the biogas was collected through a sodium hydroxide solution. Methane flow rate with a biogas
flow meter (MGC-10 PMMA, Ritter, Germany), and pH and conductivity with a multi parameter
31
sensor (pH/Cond 340i WTW, Germany) were periodically monitored. A detailed description of
biological system and procedure can be found elsewhere (Lafita et al 2015). A diagram of the
EGSB reactor and the DM contactor is shown in Fig. 1.
Fig. 1. Scheme of the set-up of the EGSB reactor and degassing membrane module
The EGSB reactor achieved a removal rate of around 31.4 kg COD m-3 d-1 resulting in organic
removal efficiency > 95%. The biogas composition was 77 % (v/v) CH4 and 23 % (v/v) CO2
and methane flow rate was 1.65 L h-1. The average D-CH4 concentration measured in the
water effluent of the reactor was around 30 mg L-1. This indicates an oversaturation of the
anaerobic effluent by a factor of 2 with respect to the equilibrium D-CH4 of 15 mg L-1 at
experimental conditions.
Membrane module was coupled to the EGSB reactor and feed with a fraction of flow from the
recirculation stream, in which the concentration of the D-CH4 was similar to the effluent stream.
The stream was flowed through a filter of 40 μm, to minimize membrane fouling, and then
through contactors, using a peristaltic pump (Watson-Marlow, USA), resulting in flow rates
from 0.36 to 10.80 L h-1. For vacuum pressure experiments, a vacuum pump N026.3.AT.18
(KNF Neuberger, Germany) was used for vacuum operation, getting vacuum pressures (Pvac)
of 140 and 500 and 800 mbar. For countercurrent sweep gas experiments, nitrogen gas with
a high purity (> 99.8 %) supplied by Carburos Metálicos S.A. (Spain) was introduced into
contactors in a range of flow rates between 2.7 and 27.0 L h-1 (STP) using a mass flow
controller (Bronkhorst Hi-Tec, The Netherlands). Liquid pressure drop at the inlet and outlet of
the membrane was measured using a portable manometer MP 112 (Kimo, Spain).
DM contactor was operated in two modes. Initially, experiments with the liquid flowing in lumen
side and vacuum pressure or sweep gas applied in the shell side (lumen side mode) were
carried out. Afterwards, the opposite operation mode was studied (shell side mode).
The determination of dissolved CH4 was carried out with a headspace method. Liquid samples
of 50 mL were collected at the inlet and the outlet of membranes and were injected in sealed
vials of 125 mL prefilled with helium. Vials were shaken vigorously for 30 seconds and left at
25 ºC for 3 hours in an orbital shaker, to allow gases to equilibrate. After equilibration, 0.5 mL
of the headspace gas was injected into a gas chromatograph (Agilent GC 7820A, Spain),
32
equipped with Agilent HP- PLOT/Q and Agilent HP-MOLESIEVE columns. The concentration
of D-CH4 in liquid phase was calculated as:
CGH (VGH + H VL )
CL = (1)
VL
where CL is the concentration of D-CH4 in liquid phase (mg L-1), CGH the concentration of CH4
in headspace after equilibration (mg L-1), VL and VGH are volumes of liquid and gas space in
the vial respectively (mL), and H the dimensionless Henry’s law constant for methane (29.55)
at 25 ºC (Sander 1999).Performance of degassing module was evaluated with the removal
efficiency (RE, %) of the membrane, defined as:
CL1 - CL2
RE = ∙ 100 (2)
CL1
where CL1 and CL2 are concentrations of D-CH4 in liquid phase (mg L-1) at the inlet and outlet
of the membrane module, respectively. Samples from inlet and outlet were taken by duplicate
and were analyzed by triplicate.
Analysis of mass transfer is important to evaluate the performance of membrane contactor

operations. Assuming negligible mass transfer resistance into gas phase, the experimental
overall mass transfer coefficient (Kexp, m s-1) can be determined from a mass balance to the
liquid phase in the contactor by the following differential expression:
dC
QL = - Kexp (CL -C*L ) (3)
dA
where QL is the liquid flow rate (m3 s-1), A is the interfacial area (m2), CL is the concentration
of CH4 in the liquid phase (mg L-1) and CL* is the equilibrium concentration of CH4 in liquid
phase (mg L-1), which can be calculated from:
C*L = H ∙CG (4)
where CG is the concentration of CH4 in the gas phase (mg L-1). Integrating expression (3)
through DM contactor:
QL CL2 - C*L
Kexp = - ln (5)
A CL1 - C*L
Cleaning of the DM module with deionized water at countercurrent flow was done after every
experiment. A control experiment was repeated every month to ensure that the DM module
was operating without fouling and the efficiency was not decreasing. Duration of every
experiment was usually 60-90 minutes, when a stable behaviour was observed. Discussion of
results has been made with values obtained at this pseudo stationary-state.
A long-time experiment (without intermediate water cleaning) was carried out at lumen side
mode with a flow rate of 1,8 L h-1, in order to determine the effect of operation time on contactor
fouling and performance. When decrease of RE was observed, a chemical cleaning was
33
carried out with 5% citric acid solution during 30 min at 1,8 L h-1 through the lumens, followed
by 5% NaOH solution at the same conditions and deionized water until pH recovery.
3.1. Vacuum pressure and sweep gas operations.

As can be seen in Fig. 2, a nearly linear increase of D-CH4 removal efficiency with the vacuum
pressure (Pvac) was observed for the whole range of liquid flow-rate studied (QL). The
variations of RE with Pvac seems to indicate that equilibrium concentration in liquid phase (CL*)
was not negligible compared with concentration in liquid phase (CL). This fact was
corroborated experimentally with CL* values (estimated from measured CG) not much lower
than concentrations in liquid phase. The increase in liquid flow-rate (QL) produced a decrease
in D-CH4 RE (Fig. 2), which could indicate that the mass transfer resistance in liquid phase
plays an important role. Similar values and behaviour of D-CH4 RE has been reported
previously at similar operational conditions (Cookney et al. 2012).
100
QL
QL==0.36
0.36 L
L h-1
h-1
90
QL
QL==66LL h-1
h-1
80 10.8 LLh-1
QL==10.80
QL h-1
70
60
RE, %
50
40
30
20
10
0
0 100 200 300 400 500 600 700 800 900
Pvac, mbar
Fig. 2. Effect of Pvac on D-CH4 RE. Lumen side mode.

At sweep gas operation, the observed behaviour was significantly different from that described
for vacuum pressure operation. As it can be observed in Fig. 3, at low values of Q N2, an
increase in the RE was achieved with an increase of QN2. Nevertheless, at high QN2 no
significant variation in the RE with sweep gas flow rate was observed. Constant RE with the
increase of gas side hydrodynamics has been also reported previously (Cookney et al. 2016).
Facts as the gas phase mass transfer resistance almost negligible and the low partial pressure
of methane in the sweep gas can explain this behavior (Bandara et al 2013), which was
corroborated experimentally since very low values of methane in gas phase were measured
(CG).
100
QL
QL==0.36
0.36L L
h-1h-1
90 QL==1.8
QL h-1h-1
1.8L L
80 QL==6 6L Lh-1h
QL
-1
70
60
RE, %
50
40
30
20
10
0
0 5 10 15 20 25 30
QN2, L h-1
Fig. 3. Effect of QN2 on D-CH4 RE. Lumen side mode.
34
The decision to operate in vacuum or sweep gas modes for the removal of D-CH4 depend on
different factors that should be taken into account, such as the D-CH4 removal efficiency, reuse
of the recovered compound and energy consumption. Considering the energy consumption,
Vallieres and Favre (2004) highlighted that for industrial scale, sweep gas operation consumed
lower raw energy than vacuum operation, except when a low vacuum pressure can be
practicable. Nevertheless, vacuum operation is preferred when it is desired to recover pure or
high purity components to reuse them as a source of energy. In this case, sweep gas operation
would involve a subsequent purification step that increases the overall energy consumption
so vacuum operation can be consider as the most feasible solution to reuse the recovered
gas.
3.2. Lumen and shell side mode operations.

As can be seen in Fig. 4a and 4b, higher RE were achieved at lumen side operation
independently of operation conditions. As Reynolds number were quite similar in both modes
of operation (Re ranging from 2.3 to 70 and from 1.3 to 40 at lumen and shell side,
respectively), the worst performance at shell side mode might be attributed to the channeling
phenomenon that can occur when liquid pass through the shell of the module, which involves
that not all fibers of the module contribute to D-CH4 RE. Channeling phenomenon was also
appointed for the removal of dissolved oxygen from water by DM contactors (Vladisavljevic
1999, Peng at al 2008). No significant effect of QN2 on performance was observed for shell
side operation.
In spite of lumen side operation has been extensively reported as more efficient, shell side
mode is usually preferred at real industrial treatments since it minimizes problems such as
high pressure drops and fouling, which can limit capacity of the modules.
100 100
a) Lumen side b) Lumen side
90 90
Shell side Shell side
80 80
70 70
60 60
RE, %
RE, %
50 50
40 40
30 30
20 20
10 10
0 0
0 2 4 6 8 10 12 0 1 2 3 4 5 6 7
QL, L h-1 QL, L h-1
Fig. 4. Effect of QL on RE at lumen and shell side operation modes. a) Pvac = 800 mbar.
b) QN2 = 27 L h-1.
3.3. Fouling
As it has been explained formerly, a long time experiment was performed in order to determine
the possible influence of fouling on the contactor performance and the frequency of cleaning
that would be used to maintain the desired capability of methane recovery/removal. As can be
seen in Fig. 5, the performance was kept almost stable/constant (RE around 60-65%) during
approximately 120 hours of operation, but after this period an appreciable and continuous
decrease in efficiency was observed. This decrease was attributed to fouling phenomena on
membrane, which was corroborated since after the chemical cleaning the initial activity was
completely recovered (around 60% of RE). The period from which membrane fouling affect
considerably the DM performance and it must be cleaned, is especially important in real
35
industrial continuous treatments, where cleaning stages must be planned, and taken into
account for economic considerations.
100
90
CLEANING
80
70
RE, % 60
50
40
30
20
10
0
0 20 40 60 80 100 120 140 160 180 200
t, h
Fig. 5. Fouling: Effect of operation time on RE. Lumen side. QL=1.8 L h-1. QN2= 12 L h-1.
3.4. Mass transfer analysis

For mass transfer equipment in general and for membrane contactors in particular, mass
transfer coefficients can be expressed using correlations of the form:
Sh = a Gz α Sc β (6)
where:
Sh = kL⋅de/D – Sherwood number,
Gz= Re⋅Pr⋅de/L – Graetz number,
Sc= νL/D – Schmidt number,
Pr=Cp⋅μ/λ – Prandtl number,
Re=vL ρ⋅de/ μ – Reynolds number,
being, de (m) hydraulic diameter, L(m) length, Cp (J/kg K) liquid specific heat, μ (kg m-1 s-1)
dynamic viscosity, ρ (kg m-3) liquid density, λ (J s-1 m-1 K-1) liquid thermal conductivity, vL (m s-
1
) liquid velocity, D molecular diffusivity of methane in water reported as 1.76 10 -9 m2 s-1
(Cengel and Boles 2008), and kL the liquid mass transfer coefficient (m s-1), which can
determined experimentally from the expression:
1 H 1 H
= + + (7)
K exp km kL kG
where kG and km are gas and membrane transfer coefficients (m s-1), respectively. Assuming
that the main resistance to mass transfer was in the liquid film resistance one can conclude
that Kexp ~ kL. In Fig. 6, the liquid phase mass transfer coefficient is plotted against the feed
water velocity in the form of Sherwood number (Sh) against Graetz (Gz) number on a log–log
scale for experiments at lumen side and vacuum pressure (500 mbar) operation. The well-
known Leveque solution (Lévêque 1928) Sh = 1.615 Gz1/3, which is the relation between
Sherwood number and Reynolds number or Graetz number, is frequently used in correlations
for lumen-feed operation. As can be seen in this figure, experimental values follow
approximately the theoretical tendency (Sh = f (vL1/3)) but slightly displaced at higher values of
36
kL. Similar results have been obtained at other operational conditions, which may indicate that
Lévêque correlation would estimate the liquid resistance of our system.
100.0
Leveque
Exp
y = 2.495x0.3415
R² = 0.9365
Sh 10.0
1.0
1 10 100
Gz
Fig. 6. Sherwood number versus Graetz number at Pvac = 500 mbar. Lumen
side operation. Dashed line: Lévêque solution.
3.5. Energy and carbon balance

As it was discussed formerly, vacuum operation offers the potential to deliver much higher
recovered gas purities, since it minimised gas-side dilution by avoiding the addition of dilutive
sweep gas, and an elevated concentration of methane can be obtained. Therefore, only this
operation mode has been evaluated for the energy and cost efficiency.
In order to determine the efficiency of the recovery of methane, a comparison between energy
recovered as methane and energy demand of the operation must be carried, in this case
power demand of vacuum (blower) and liquid flow (pump) operations. Blower/compressor
power consumption (W) can be calculated for an isentropic process from the following
expression:
γ
γ R Ta PD γ−1
(8)
W(J kg −1 )= −1
γ−1 M PA
PD and PA are discharge and suction pressure, respectively, γ is isentropic expansion

coefficient of gas, R is the constant of ideal gases, M is the molecular weight of the gas, T a is
the temperature of the suction. Assuming a vacuum pump efficiency of 0.65, the power
demand can be calculated. For liquid flow, the power consumption in membrane can be
estimated from:
PI − PO (9)
W(J kg−1 )= + ∑F
ρ
where PI and PO are inlet and outlet pressure, ∑F is the friction loss (estimated as a 20% of
the total power demand) and ρ is the liquid density. A pump efficiency of 0.65 was also
estimated.
37
Assuming an electrical conversion efficiency of methane of 35%, the energy required in

vacuum operation ranged from 0.2% to 3% of the recovered energy in membrane as methane
for 140 mbar and 800 mbar of vacuum pressure, respectively. The pump operation was always
lower than 0.01% of the recovered energy and therefore it can be considered as negligible.
As the amount of the recover methane increased with vacuum pressure and RE, to compare
different operational conditions the parameter Saved Energy has been defined as the
difference between the recovered electrical and demand of vacuum operation energies per
cubic meter of treated effluent.
800 mbar LS 500 mbar LS 140 mbar LS 800 mbar SS
600
Saved energy (J m3 effluent)
500
400
300
200
100
0
0 0.05 0.1 0.15 0.2 0.25 0.3
Anaerobic effluent flow-rate (m3 day-1)
Fig. 7. Recovered energy at different operational conditions. Lumen
(LS) and shell side (SS) mode operations.
As is shown in Fig 7, the saved energy was positive in all cases, and an optimum value of
liquid flow rate for lumen side operation around 0.05-0.15 m3 day-1 was found. These results
show that a net energy production can be obtained from the recovery of methane in anaerobic
effluents. In addition, the renewable energy that can be produced from the recovered methane
can offset the carbon footprint of D-CH4 in anaerobic effluents, which indicate that integrating
this technology in the anaerobic treatment of wastewater even can become carbon positive.
The environmental interest of this technology is clearly demonstrated when the emission to
the atmosphere of around 1 kg CO2 equivalent m-3effluent is avoided every 15% of removal/recovery
efficiency (RE) of the process. In a conventional anaerobic treatment of 10 m3effluent h-1 and for
a RE of 60-70%, 350-450 tons of CO2 equivalent per year are not emitted.
From the point of view of economic viability, the energy saved seems relatively low: for
example, 500 J m-3effluent (near the maximum value in Fig. 7) means that approximately 0.14
kWh m-3effluent can be saved and assuming an electricity cost of 0.100 € kWh-1, only around
0.014 € per cubic meter of treated effluent could be obtained in the methane/energy recovery
operation. In this conditions, for a conventional anaerobic treatment of 10 m3effluent h-1, the
economic save from produced electricity would be of only around 1000-1200 € per year.
Taking into account that investment, personnel and other costs might be supply by this save,
the economic viability of this technology is compromised.
38
4. CONCLUSIONS
 Hollow fiber membrane contactor module has been successfully demonstrated for
dissolved methane recovery from anaerobic effluents.
 Methane removal efficiencies ranging from 25% to 75 % at experimental conditions
can be achieved, showing the viability of the methane removal/recover using this
technology.
 Lumen mode operation was more efficient.
 Results obtained at vacuum pressure experiments showed that concentration of
methane specie in the gas phase is not enough low to consider that equilibrium
concentration in liquid phase was negligible compared to actual concentration in liquid
phase (CL >>CL*; CL - CL* ≈ CL). The opposite was observed in sweep gas operation,
where the influence of nitrogen flow rate was negligible, from intermediate nitrogen
flow rates, showing that the gas phase mass transfer resistance was almost negligible.
 Mass transfer liquid phase seems to control the process.
 After 100-120 hours of operation fouling was observed and it was necessary cleaning
the contactor to recover the former RE.
 Dissolved methane recovery of anaerobic effluents showed a net energy balance and
therefore a considerable reduction of carbon equivalent emissions.
ACKNOWLEDGMENTS
Financial support was obtained from the Ministerio de Economía y Competitividad (Spain,
Project CTM-2014-54517-R co-financed with FEDER funds). Financial support was also
obtained from Generalitat Valenciana (Spain, Prometeo project 2013/53). M. Henares
acknowledges Generalitat Valenciana for financing support of her PhD.
REFERENCES
Bandara WMKRTW, Ikeda M, Satoh H, Sasakawa M, Nakahara Y, Takahashi M, Okabe S

(2013) Introduction of a Degassing Membrane Technology into Anaerobic Wastewater
Treatment. Water Environ. Res. 85:387–390.
Cakir FY, Stenstrom MK (2005) Greenhouse gas production: a comparison between aerobic
and anaerobic wastewater treatment technology. Water Res. 39: 4197–203.
Cengel YA, Boles MA (2008) Thermodynamics: An Engineering Approach, 6th Edition.
McGraw-Hill, New York.
Cookney J, Cartmell E, Jefferson B, McAdam EJ (2012) Recovery of methane from anaerobic
process effluent using poly-di-methyl-siloxane membrane contactors. Water Sci. Technol.
65:604–610.
Cookney J, Mcleod A, Mathioudakis V, Ncube P, Soares A, Jefferson B, McAdam EJ (2016)
Dissolved methane recovery from anaerobic effluents using hollow fibre membrane
contactors, J. Memb. Sci. 502:141–150.
El-Fadel M, Massoud M (2001) Methane emissions from wastewater management. Environ.
Pollut. 114:177–185.
Ito A, Yamagiwa K, Tamura M, Furusawa M (1998) Removal of dissolved oxygen using non-
porous hollow-fiber membranes. J. Memb. Sci. 145:111–117.
Lafita C, Penya-roja JM, Gabaldón C (2015). Anaerobic removal of 1-methoxy-2-propanol
under ambient temperature in an EGSB reactor. Bioprocess Biosyst. Eng. 38(11):2137–
2146.
39
Lévêque A (1928) Les lois de la transmission de chaleur par convection. Annales des
Mines,12-13:201–299.
Leiknes T (2001) Vacuum degassing using microporous hollow fiber membranes. Sep. Purif.
Technol. 22-23:287–294
Lettinga G, Rebac S, Zeeman G (2001) Challenge of psychrophilic anaerobic wastewater
treatment. Trend in Biotechnology 19 (9): 363-370.
Peng ZG, Lee SH, Zhou T, Shieh JJ, Chung TS, A study on pilot-scale degassing by
polypropylene (PP) hollow fiber membrane contactors (2008) Desalination 234: 316–322.
Sander R (1999) Compilation of Henry’s Law Constants for Inorganic and Organic Species of
Potential Importance in Environmental Chemistry. Database 20:107.
Vallieres C, Favre E (2004) Vacuum versus sweeping gas operation for binary mixtures
separation by dense membrane processes. J. Memb. Sci. 244:7–23
Vladisavljevic G (1999) Use of polysulfone hollow fibers for bubbleless membrane
oxygenation/ deoxygenation of water. Sep. Purif. Technol. 17:1–10.
40
Oral Presentations – Residues Valorization I

41
Assessing the potential of industrial waste streams

for volatile fatty acids (VFA) production
Garcia-Aguirre, Jon1, Aymerich, Enrique1, and Esteban-Gutiérrez, Myriam1
1 Ceit-IK4 and Tecnun (University of Navarra), 15 Paseo Manuel de Lardizabal, San Sebastian 20018,
Spain (E – mail: jgaguirre@ceit.es; eaymerich@ceit.es; mesteban@ceit.es)
ABSTRACT
The acidogenic potential of six different agro industrial waste streams was assessed for
volatile fatty acids (VFA) production. The total VFA profile was obtained and the VFA
composition profile was analyzed at acidic and alkaline pH. Higher net VFA production was
observed at alkaline conditions suggesting that the hydrolysis and solubilization of substrates
were enhanced. The VFA yield ranged within 162–273 mg VFA (COD) g-1 CODfed and 283–
811 mg VFA (COD) g-1 VSfed with a degree of acidification (DA) of 12–29% at alkaline pH.
Crude Glycerol presented the highest VFA concentration peak and the greatest VFA yield
after 10 days of fermentation. The results obtained evidenced that the composition of the
substrate and the fermentation pH exert a remarkable effect on the final product distribution.
Besides, the information obtained contributes to add knowledge on VFA production from agro
industrial by-products and it is valuable as a first stage to assess the feasibility of on–site VFA
production for the valorization of agro industrial resources.
Keywords: VFA; Acidogenic potential; Waste Valorisation; Industrial Waste Streams.
1. INTRODUCTION
Modifications in environmental legislation such as the entry into force of the Landfill Directive
1999/31/EC and the Waste Framework Directive 2008/98/EC, in addition to the increasing
social and environmental concern on the pollution of the planet, have been the main drivers
in directing the perception of organic waste towards a valuable resource. More recently, the
EU has published the “Closing the loop - An EU action plan for the Circular Economy” (COM
(2015) 614), which includes specific measures from production and consumption to waste
management, and proposes concrete measures to promote re-use and stimulate industrial
symbiosis - turning one industry's by-product into another industry's raw material.
Anaerobic digestion (AD) has been widely studied as an energy–efficient and environmentally
friendly technology for bioenergy production from organic wastes. Despite the fact that AD
may be a mature technology, commonly economic subsidies are needed to overcome the
competition with natural gas (Kleerebezem et al., 2015). Dark fermentation (DF) has also
attracted research attention as a promising way for biohydrogen production (Ghimire et al.,
2015). Nonetheless, due the low hydrogen yield achieved in the process, its extrapolation to
large scale applications has not been carried out yet.
Volatile fatty acids are valuable intermediate chemicals of the aforementioned biodegradation
processes. VFA can be used in a wide range of applications such as cosmetics production,
petrochemical synthesis, in the pharmaceutical industry or in the food and beverage industry
(Zacharof and Lovitt, 2013). Although its chemical synthesis is typically performed by using
non-renewable feedstock, its production via mixed culture fermentation has open an
opportunity for novel applications such as biodegradable polymer production (PHA)
(Megmeng et al., 2009, Tamis et al., 2014, Shen et al., 2014), nutrient removal in wastewater
treatment plants (WWTP) (Lim et al., 2008, Zheng et al., 2010), medium chain fatty acid
production (Grootscholten et al., 2013), bioenergy production in microbial fuel cells (MFC)
(Pant et al., 2010, Cavdar et al., 2011) or bulk fuel and solvent production (Agler et al., 2011).
Indeed, a new bio-refinery concept has emerged based on VFA production by means of mixed
culture engineered hydrolysis and fermentation processes (Chang et al., 2010).
42
Nonetheless, VFA production from acidogenic fermentation is still in its early stage.
Operational parameters such as the temperature, pH, the retention time and the organic
loading rate exert a direct influence on the VFA production rate and yield (Lee et al., 2014).
These factors determine which reactions are thermodynamically feasible and which
microorganisms are involved in the reactions, thus, influencing the final product spectrum.
Further research on the control of the process may help control the products formed during
the acidogenic fermentation (Kleerebezem et al., 2015). Currently, effectively inhibiting the
hydrogenotrophic methanogenesis, directing the microbial process to generate desired
products or achieving an effective separation of VFA from the fermentation broth remain as
major challenges for large scale VFA production (Agler et al., 2011).
The commercial VFA production process should use mixed microbial cultures to handle the
diverse nature of different organic waste streams (Agler et al., 2011). Depending on the readily
fermentable organic matter, different substrates may be adequate for VFA production.
Choosing a specific waste stream will depend on the locally available resources and in the
scope of application. While the organic fraction of municipal solid waste (OFMSW) and
sewage sludge have been a focus of study during recent years (Ucisick and Henze, 2008, Lim
et al., 2008, Xiong et a., 2012, Jiang et al., 2013, Pittman and Steinmetz, 2013), there is limited
information available on VFA production from agro industrial resources. Indeed, assessing
and comparing the acidogenic potential of different organic wastes and optimizing the process
parameters may be valuable for different industrial areas.
In this study, the acidogenic potential of six different agro industrial waste streams was
assessed which included wastewater streams and sludges. Fermentation tests were
performed in laboratory scale batch experiments at acidic (pH 5.5) and alkaline (pH 10.5)
conditions. By means of comparing different substrates, the VFA production profile, the VFA
yield and the VFA composition profile were evaluated.
2.1. Organic waste streams

Different agro industrial waste streams were assessed for VFA production. These consisted
of three organic effluents: slaughterhouse wastewater (Sww), paper mill wastewater (Pww) and
winery wastewater (W ww); two organic sludges, i.e. waste activated sludge (WAS) and dairy
sludge (DS); and crude glycerol (CG).
Waste activated sludge was obtained from the thickener of a municipal WWTP. Dairy sludge
was taken from the WWTP of a milk production factory. A Kraft paper mill factory and a local
slaughterhouse provided their respective wastewaters. The winery wastewater was obtained
from the effluent discharged in a winery waste valorisation plant. Crude glycerol was obtained
from a biodiesel production plant. The main characteristics of the substrates used in the
present study are shown in Table 1.
Digested sludge was collected, from the anaerobic digestor of the above mentioned municipal
WWTP before the beginning of each experiment and it was used as inoculum. The inoculum
was collected before each experimental assay to avoid its degradation during long time
storage. All waste samples and inoculum were stored at 5ºC prior to use.
43
Table 1. Characteristics of the substrates used in this study.
Parameter WAS CG DS Sww Pww W ww

TS (%) 6.3 17.6 6.4 0.41 0.60 9.3
VS (%) 4.8 7.9 5.5 0.34 0.26 7.0
VS/TS 0.77 0.45 0.86 0.82 0.43 0.76
TSS 5.7 0.63 5.8 0.24 0.038 5.5
VSS 4.7 0.13 5.5 0.23 0.035 5.0
tCOD (mg O2 L-1) 72445 238873 143612 6090 5274 101085
sCOD 2373 233791 18852 1502 4265 30337
TKN (mg L-1) 451 340 3881 209 63 3612
TAN (mg L-1) 109 169 952 22 8 112
pH 6.2 4.4 5.7 7.6 6.0 5.1
BA (mg L-1) 1120 0 3595 307 20 1940
TA (mg L-1) 510 75 70 205 1110 0.0
Proteins (mg DQO L-1) 3032 1516 25993 1655 491 31062
Lipids (mg DQO L-1) 2388 1009 8586 375 N.D 741
2.2. Batch fermentation tests

The acidogenic fermentation tests were performed in duplicate, in batch assays and 500 mL
bottles with a working volume of 400 mL. Both acidic (pH 5.5) and alkaline (pH 10.5) conditions
were assessed for each substrate, with blank tests containing solely inoculum as control.
Based on previous studies, 10 fermentation days were used to assess the VFA production of
the different waste streams.
In each reactor the volume of substrate was determined and adjusted to an initial COD
concentration of 10000 mg L-1, except for the Sww and Pww, where the initial COD was adjusted
to 4000 mg L-1. Similarly, the volume of inoculum was determined to set a biomass
concentration of 5000 mg SSV L-1 and 2000 mg SSV L-1 for Sww and Pww. Neither nutrient
solutions nor methanogenic inhibitors were used. The pH was initially adjusted to 5.5 and 10.5,
by adding HCl and NaOH respectively, to avoid the methanogenic activity of microorganisms.
Distilled water was added to make up the final volume.
Nitrogen gas was flushed to the bottles for 1 min to remove any residual oxygen and the bottles
were then sealed with rubber stoppers. The reactors were placed in a thermostatic incubator
at mesophilic conditions (35ºC) for 10 days. Mixing was performed manually twice/once a day.
At the beginning of the tests, a full characterization of the substrates (waste streams) and of
the initial mixture was performed. 3–5 mL of sample was extracted from the reactors on a daily
basis to monitor the VFA production and the pH evolution. Biogas production and composition
were monitored daily. A final characterisation of the bottles was conducted to analyse the
composition of the fermentation mixtures after a 10 day incubation period.
2.3. Analytical methods

Total solids (TS), volatile solids (VS), total suspended solids (TSS), volatile suspended solids
(VSS), total chemical oxygen demand (tCOD) and soluble chemical oxygen demand (sCOD),
Total ammonia nitrogen content (TAN) and total Kjeldahl nitrogen (TKN), bicarbonate alkalinity
(BA), total alkalinity (TA) and pH were determined following the guidelines given by the
standard methods 20th edition (1998).
Biogas production was monitored daily and the composition analyses (GC-TCD HP6890,
column SUPELCO 60/80 Carboxen, Ref. 10001-2390-U) were carried out on a daily basis in
order to check the methane production.
VFA concentration (acetic acid, propionic acid, isobutyric acid, butyric acid, isovaleric acid and
valeric acid) was determined with an Agilent GC–6890 gas chromatograph equipped with a
44
FID (Flame Ionization Detector) and a capillary column (DB-FFAP, 30m x 0.25 mm i.d., 0.25
µm film, Agilent J&W: ref. 122-3232E). The filtered fraction was obtained by means of
centrifugation (4000 rpm, 5-10 min) followed by vacuum filtration (Whatman 1.5 µm) and it
was processed by liquid – liquid extraction with an organic solvent (Ter-Butil-Metyl-Eter) in
order to obtain the sample for VFA determination. Pivalic acid was used as an internal
standard solution.
2.4. Calculations
VFA yield (i.e. total VFA production per substrate) was calculated through the sum of
all the individual VFA at the maximum acidic production, by extracting the initial VFA content
of the substrate. The VFA produced by the inoculum was excluded from the total VFA
production. The Degree of acidification (DA) was defined as the total VFA produced
exclusively by the substrate by the influent’s total COD (Silva et al., 2013). For a
straightforward comparison between substrates, the VFA yield in terms of VFA (in COD units)
g-1 CODfed to the reactors and the amount of VFA (in COD units) g-1 VSfed was calculated.
3.1. VFA production profiles

Figure 1 represents the total VFA expressed as COD equivalents L-1 of the substrates tested
during the fermentation period at both acidic (pH 5.5) and alkaline (pH 10.5) conditions.The
initial pH adjustment enabled no biogas production at alkaline conditions while some biogas
with low methane content could be observed at acidic conditions. Indeed, other researchers
have used methanogenic inhibitors to suppress the activity of methanogenic microorganisms.
At the end of the anaerobic fermentation period, the maximum VFA production had been
reached at acidic conditions for all experiments, although CG and W ww might still show some
residual acidogenic potential. This was similar at alkaline conditions, were CG and W ww
showed an increasing VFA production trend on the 10th day.
Figure 1. a) Total VFA production at acidic conditions. b) Total VFA production at alkaline conditions.
Overall, the fermentation mixtures were acidified although the fermentation pattern was
different depending on the substrate considered. Under acidic conditions, the substrates
tended to acidify from the beginning although a lag phase was observed for Pww, where the
mixture was not acidified until day 7 (Figure 1a). In addition, the acidification of WAS was
almost negligible at acidic conditions, which was likely due to the nature of the substrate which
consists of complex polymeric materials (Tyagi and Lo, 2013). Under alkaline conditions, an
initial lag phase was also observed for CG, W ww and Pww, which suggests that the biomass
needed an acclimation period to the substrate (Figure 1b). Regarding the results of DS, the
VFA were initially consumed at both pH conditions.
45
As shown in Figure 1, the VFA concentration was higher at alkaline conditions for all
substrates. CG presented the highest VFA concentration, with a peak of 4257 mg VFA (COD)
L-1 on the 10th day of fermentation, followed by WAS, W ww and Pww, with 3185, 2745 (9th day)
and 2511 mg VFA (COD) L-1, respectively. Similarly, at acidic conditions, CG, W ww and Pww
achieved the highest VFA concentrations with 3055, 2583 (9th day) and 2281 mg VFA L-1 as
COD. Lower VFA values were produced by DS and Sww, which were in the range 700-1234 at
acidic conditions and 794-1615 mg VFA (COD) L-1 at alkaline conditions.
3.2. VFA yield

Table 2 represents the VFA yields and the degree of acidification (DA) of the substrates that
have been analyzed in the present study. Since in WAS mainly all VS are in the form of VSS,
the VFA yield in terms of mg VFA (COD) g-1 VSfed was considered so that a comparison
between different substrates could be made. The VFA yields ranged within 162–273 mg VFA
(COD) g-1 CODfed and 283–811 mg VFA (COD) g-1 VSfed at alkaline conditions. Crude glycerol
showed the highest VFA potential followed by Pww in terms of mg VFA (COD) g-1 VSfed, which
is likely due to the soluble nature of both substrates, which are already hydrolyzed. Overall,
the VFA yields were lower at acidic conditions, especially for WAS which showed almost no
VFA production. Previous literature studies reported that alkaline conditions favour VFA
production from WAS as the hydrolysis and solubilization of substrates and thus the
subsequent VFA production was favoured (Chen et al., 2007, Yu et al., 2008, Zhang et al.,
2009). The final concentration of VFA achieved in this study, namely 3185 mg COD L-1 and
the VFA yield of 283 mg VFA (COD) g-1 VSfed, are within the values reported in previous
studies. Moreover, it is likely that the initial pH adjustment acted as a pre-treatment step for all
substrates, by increasing the solubility and the hydrolysis and the VFA production as
suggested by Lee et al. (2014).
Table 2. VFA yield achieved during 10 days of fermentation and the degree of acidification.
mg VFA(COD) g-1 mg VFA(COD) g-1 DA
Substrate CODfed VSfed
pH 5.5 pH 10.5 pH 5.5 pH 10.5 pH 5.5 pH 10.5
Sww 140 162 250 290 8.1 12.4
Pww 143 182 293 372 10.1 23.3
CG 182 273 540 811 12.8 28.6
W ww 190 207 273 297 11.6 16.2
WAS 0.9 190 1.4 283 0.14 18.6
A higher DA was achieved at alkaline pH, falling within 12-29%. The DA measured for CG
(12.8%) at acidic conditions was similar to the results reported by Silva et al. (2013), since
13% DA was reported for crude glycerol at mesophilic conditions. At alkaline conditions, the
DA (29%) was similar to the value reported by (Forrest et al., 2010), who run a study at
thermophilic conditions. The DA of W ww at alkaline was equal to the value reported by Silva et
al. (2013) of 11 %, which was enhanced at alkaline conditions. The highest DA of 26 % of
COD was reached with dairy fatty slurry with an initial load of 15 g COD L-1 (Arroja et al., 2012).
3.3. Composition of the fermentation product

Overall, under acidic conditions the predominant fermentation products were acetic, propionic
and butyric acids, which are common primary fermentation products (Agler et al., 2011) while
iso-butyric, n-valeric and iso-valeric acids appeared in a lower extent (Figure 2).
46
Figure 2. Fermentation products after 10 day incubation: a) Individual VFA production, and b)
composition of the fermentation broth.
As shown in Fig. 2 and Fig. 3, different substrates led to different fermentation products. At
acidic pH, the main fermentation product was propionic acid, with 2271 mg VFA (COD) L-1,
which accounted for 74% of the fermentation mixture. Zhang and Yang (2009) obtained high
concentrations of propionic acid by metabollically engineered Propionibacteria. Kosmider
(2010) highlighted the possiblity of using crude glycerol for propionic acid production. Silva et
al. (2013) reported a higher proportion of acetic than butyric acid during CG fermentation.
However, the pH values remained higher than 7, while in the present study the pH values were
kept close to pH 5.5. This suggests that the pH could be controlled with CG to exert influence
in the final product distribution. Indeed, acetic acid was favoured (91%) in our study at alkaline
conditions (Fig. 3).
Butyric acid production was promoted in sigfinicat amounts with Sww and DS (23– 24%,
respectively), while iso–valeric acid accounted for 32% of total VFA, although the overall VFA
production and acidification was minimum. A similar VFA composition was obtained by Arroja
et al. (2012), who reported acetic acid and propionic acid as the main fermentation products
from dairy fatty slurry. Pww fermentation produced acetic and propionic acids and lower
amounts of butyric acid. Bengtsson (2008) also reported these acids as the main components
during the acidogenic fermentation of different paper mill wastewaters at pH 6. Butyrate,
acetate, ethanol and propionate were identified as the major fermentation products during the
batch tets performed by Jiang (2012) at pH 6 and 15–25 ºC. All this evidences the compositon
of the substrate exerts a remarkable effect on the final product composition.
Overall, the composition of the fermentation mixtures changed at alkaline conditions, as shown
in Figure 3. This corroborates previous findings, which suggested that the pH exerts a major
effect in the final acid distribution (Lee et al, 2014). Acetic acid was the predominant
fermentation product for all substrates, ranging from 43 to 91%. In this study, longer chain
fermentation products (butyric, iso–butyric, valeric and iso–valeric acids) were present in
significant amounts, accounting for 27–35% during the fermentation of Sww, DS and WAS.
Propionic acid production was lower for all substrates compared with the acidic tests, except
for Pww, where a 39% of the total fermentation broth was observed.
47
Figure 3. Fermentation products after 10 day incubation: a) Individual VFA production.

b) Composition of the fermentation broth.
The individual VFA composition of Sww and DS was almost equal, despite the differences in
the VFA production trend. This may be due to the similar composition of both substrates, which
were rich in proteins and lipids (Table 1).
Although the pH and the type of substrate exerted a clear effect in the final product distribution,
further research may be necessary in order to elucidate whether a selective production of
specific acids may be possible, by controlling the process parameters and by means of a
better understanding of the biochemistry of the process. With some substrates, such as CG,
shifting the pH could offer a means for controlling the final product distribution, as previously
suggested (Horiouchi et al, 1999). However, it remains unclear how other substrates interact
with the microorganisms present in the biomass or the methabolic pathways that are dominant
in the fermentation of substrates by mixed culture fermentations.
4. CONCLUSIONS
 Monitoring the daily VFA production and the assessment of the VFA yield enabled to
compare six different organic wastes in terms of VFA production potential. The results
of the study may be valuable for industry as a first assessment of the acidogenic
potential of their waste streams.
 The results reported here showed higher net VFA productions at alkaline conditions
what suggests that the hydrolysis and solubilization of substrates were enhanced.
Overall the VFA yield ranged within 162–273mg VFA (COD) g-1 CODfed and 283–811
mg VFA (COD) g-1 VSfed at alkaline conditions.
 The highest VFA potential was achieved for CG where the highest VFA concentration
and DA (29%) were obtained. The final product distribution changed with CG from
propionic to acetic acid production by shifting the pH from acidic to alkaline conditions.
 The composition of the substrate exerted a clear effect in the final fermentation product
distribution as the individual components interact differently with the biomass of the
inoculum. Thus, depending on the scope of application efforts could be made to control
the composition of the fermentation broth. Besides, VFA production might improve by
mixing substrates with complementary characteristics, where synergies between
substrates enhance the overall VFA production.
48
REFERENCES
Agler, M.T., Wrenn, B.A., Zinder, S.H., Angenent, L.T. (2011). Waste to bioproduct conversion with
undefined mixed cultures: the carboxylate platform. Trends in Biotechnology 29-2.
Arroja, L., Capela, I., Nadais, H., Serafim, L.S. Silva, F. (2012). Acidogenic valorisation of high strength
waste products from food industry industrial waste. ISBN: 978-953-51-0253-3, In tech, available at:
http://www.intechopen.com/books.
Bengtsson, S., Hallquist, J., Werker, A., Welander, T. (2008). Acidogenic fermentation of industrial
wastewaters: Effects of chemostat retention tiem and pH on volatile fatty acids production.
Biochemical Engineering Journal 40:492-499.
Cavdar, P., Yilmaz, E., Tugtas, A.E., Calli, B. (2011). Acidogenic fermentation of municipal solid waste
and its application to bio – electricity production via microbial fuel cells (MFCs). Water Science &
Technology 64.4.
Chang, H.N., Kim, N-J., Kang, Jeong, C.M. (2010). Biomass – derived Volatile fatty acid platform for
fuels and chemicals. Biotechnology and Bioprocess Engineering 15:1–10.
Chen, Y., Jiang, S., Yuan, H., Zhou, Q., Gu, G. (2007). Hydrolysis and acidification of waste activated
sludge at different pH. Water Research 41:683-689.
Forrest, A.K., Sierra, R., Holtzapple, M.T. (2010). Effect of glycerol type and fermentor configuration on
mixed-acid fermentations. Bioresource Technology 101(23):9185-9189.
Ghimire, A., Frunzo, L., Pirozzi, F., Trably, E, Escudie, R., Lens, P.N.L, Esposito, G. (2015). A review
on dark fermentative biohydrogen production from organic biomass: process parameters and use of
by – products. Applied Energy 144:73–95.
Grootscholten, T.I.M., Steinbuch, K.J.J, Hamelers, H.V.M., Buisman, C.J.N. (2013). Improving medium
chain acid productivity using chain elongation by reducing the hydraulic retention time in an upflow
anaerobic filter. Bioresource technology 136: 735-738.
Horiuchi, J., Simizu, T., Kanno, T., Kobayashi, M. (1999). Dynamic behaviour in response to pH shift
during anaerobic acidogenesis with a chemostat culture. Biotechnology techniques 13:155-157.
Jiang, Y., Marang, L., Tamis, J., van – Loosdrecht, M.C.M., Dijkman, H., Kleerebezem, R. (2012). Waste
to resource: Converting paper mill wastewater to bioplastic. Water Research 46:5517-5530.
Jiang,J., Zhang, Y., Li, Kl., Wang, Q., Gong, C., Li, M. (2013). Volatile fatty acids production from food
waste: effects of pH, T and OLR. Bioresource Technology 143:525-530.
Kleerebezem, R., Josse, B., Rozendal, R., van Loosdrecht M.C.M. (2015). Anaerobic digestion without
biogas?. Rev Environ Sci Biotechnol. Springerlink.com.
Kosmider, Alicja, Drozdzynska, A., Blaszka, K., Leja, K., Czaczyk, K. (2010). Propionic acid production
by Propionibacterium freudenreichii ssp. Sheermani using crude glycerol and whey lactose industrial
wastes. Polish J. of Environ. Stud. 19(6):1249-1253.
Lee, W.S., Chua, A.S.M., Yeoh, H, K., Ngoh, G.C. (2014). A review of the production and applications
of waste – derived volatile fatty acids. Chemical Engineering Journal 235:83-99.
Lim, S.J., Kim, B.J., Jeong, C-M., Choi, J-d-r, Ahn, Y.H., Chang, H.N. (2008). Anaerobic organic acid
production of food waste in once-a-day feeding and drawing-off bioreactor. Bioresource Technology
99:7866-7874.
Megmeng, C., Hong, C., Quingliang, Z., Shirley, S.N., Jie, R. (2009). Optimal production of
polyhydroxyalkanoates (PHA) in activated sludge fed by volatile fatty acids (VFAs) generated from
alkaline excess sludge fermentation. Bioresource Technology 100:1399-1405.
Pant, D., van Bogaert, G., Diels, L., Vanbroekhoven, K. (2010). A review of the substrates used in
microbial fuel cells (MFCs) for sustainable energy production. Bioresource technology 101:1533-
1543.
Pittman, T., Steinmetz, H. (2013). Influence in operating conditions for volatile fatty acids enrichment
as a first step for polyhydroxyalkanoate production of municipal wastewater treatment plant.
Bioresource Technology 148:270-276.
Silva, F.C., Serafim, L.S., Nadais, H., Arroja, L., Capela, I. (2013). Acidogenic fermentation towards
valorisation of organic waste streams into volatile fatty acids. Chemical Biochemical Engineering
Quarterly, 27(4):467-476.
Shen, L., Hu, H., Ji, H., Cai, J., He, N., Li, Q., Wang, Y. (2014). Production of poly(hydroxybutyrate –
hydroxyvalerate) from waste organics by the two stage process: focus on the intermediate volatile
fatty acids. Bioresource Technology 166:194-200.
Tamis, J., Luzkov, K., Jiang, Y., van Loodrecht, M.C.M., Kleerebezem, R. (2014). Enrichment of
Plasticicumulans acidivorans at pilot scale for PHA production on industrial wastewaters. Journal of
Biotechnology 192:161-169.
Tyagi, V.K., Lo, S-L. (2013). Sludge: a waste or renewable source for energy and resources recovery?.
Renewable and Sustainable Energy Reviews 25:708-728.
49
Ucisik, A.S., Henze, M. (2008). Biological hydrolysis and acidification of sludge under anaerobic
conditions: the effect of sludge type and origin on the production and composition of volatile fatty
acids. Water research 42, 3729–3738.
Yu, G-H., He, P.J., Shao, L-M, He, P.P. (2008). Towards understanding the mechanisms of improving
the production of volatile fatty acids from activated sludge at pH 10. Water Research 42:4637-4644.
Xiong, H., Chen, J., Wang, H., Shi, H. (2012). Influences of volatile solids concentration, temperature
and solid retention time for the hydrolysis of waste activated sludge to recover volatile fatty acids.
Bioresource Technology 119:285-292.
Zacharof, M.P., Lovitt, R.W. (2013). Complex effluent streams as a potential source of volatile fatty
acids. Waste Biomass Valor. 4:557-581.
Zhang, P., Chen, Y., Zhou, Q. (2009). Waste activated sludge hydrolysis and short – chain fatty acids
accumulation under mesophilic and thermophilic conditions: effect of pH. Water Research 43, 3735-
3742.
Zheng, X., Chen, Y., Liu, C. (2010). Waste activated sludge alkaline fermentation liquid as carbon
source for biological nutrients removal in anaerobic followed by alternating aerobic–anoxic
sequencing batch reactors. Biotechnology and Bioengineering 18(3):478-485.
50
Crude glycerol use as carbon source in a sulfate-

reducing UASB for sulfur recovery from S-rich
effluents
Mabel Mora1*, Javier Lafuente1, Jordi Fernández2, Ricard March2, David Gabriel1
1GENOCOV Research Group, Department of Chemical, Biological and Environmental Engineering,
Escola d’enginyeria, Universitat Autònoma de Barcelona, 08193 Bellaterra, Spain.
2Nubiola Spain (FERRO Corporation), Pablo Iglesias 98-100, 08908 L’Hospitalet de Llobregat, Spain
(*corresponding author: mariaisabel.mora@uab.cat)
ABSTRACT
In addition to a range of S-rich effluents from industrial wastewaters, exhaust flue gases
obtained from fossil fuels combustion or other industrial activities may contain large amounts
of SOx which must be removed before their emission to the atmosphere because of their
negative effect on human health, livestock and plants among others. After chemical absorption
of SOx, a sulfite/sulfate rich effluent is produced. Efficient but expensive physical and chemical
treatments for flue gas desulfurization such as wet or dry scrubbing require of further
expensive management or treatment. Thus, biological recovery of sulfur from such S-rich
effluents becomes an interesting alternative. Since no biological mechanism has been
described for the direct reduction of sulfite/sulfate to elemental sulfur, potential sulfur recovery
techniques require the design of a multiple-step process with H2S as intermediate product
produced through the biological reduction of SO42- to S2- under anaerobic conditions. In this
work, crude glycerol was selected as potential electron donor and carbon source for sulfate-
reducing bacteria once compared its sulfate-reducing capacity with pig slurry, cheese whey
and vinasses in batch, sealed bottles. Later, an upflow anaerobic sludge blanket reactor fed
with crude glycerol was operated and monitored for 4 months at a S-loading rate of 113 g S
m-3 h-1 and variable organic loading rates corresponding to S/C ratios from 0.8 to 2.8 g C g S-
1
to assess the startup and sulfate reduction capacity. Sulfate reduction capacities up to 110
g S m-3 h-1 were reached at the highest C/S ratio tested. In addition to biogas production, the
granular sludge bed stratification and the biological processes occurring at different heights of
the reactor were assessed during reactor operation.
Keywords: Crude glycerol, flue gases, S recovery, SOX, sulfate reducing bacteria.
1. INTRODUCTION
Combustion of sulfur-containing fuels results in SO2 formation which cause detrimental
impacts on human health and the environment such as respiratory illness and acid rain,
respectively, among others (Srivastava & Jozewicz 2001). In the last decades, SO2 release to
the atmosphere in Europe has been limited by different directives leading to a decrease in its
emission by around 53% in the last decade. However, anthropogenic worldwide emissions of
SO2 in 2011 were around 100 Tg mainly generated in the energetic and industrial sector
(Klimont et al. 2013). The treatment of SO2 emissions from flue gases is usually performed
through physical-chemical treatments, which are expensive and generate effluents that
require further processing as limestone forced oxidation or furnace sorbent injection
(Srivastava & Jozewicz 2001; Philip & Deshusses 2003). Biological processes offer
advantages with respect to physical-chemical processes mainly related with operational costs
and wastes generation. Moreover, biological processes offer the possibility of waste
51
valorization as occurs with biogas generation from anaerobic digestion (Scherson & Criddle
2014).
Wet flue gas desulfurization with sodium hydroxide is one of the most applied physical-
chemical treatment which generates aqueous slurries with a high content in sulfite and sulfate.
The development of a biotechnological process to valorize this type of liquid effluents in an
economical, robust and environmentally friendly way is challenging since could be applied not
only to valorize SO2 from flue gases but also S-rich liquid effluents in general. Previous works
have reported the reduction of sulfate to hydrogen sulfide as a previous step required for
further S and metals recovery from mining effluents processing (Gomez et al. 2010; Papirio et
al. 2013). Alternatively, sulfide-oxidizing bacteria could be used for the partial oxidation of
hydrogen sulfide to elemental sulfur for S recovery. In all cases, a highly robust, economical
and efficient biological sulfate-reducing bioreactor is needed for process competitiveness.
Other authors have proposed alternatives for valorization of SO2 from flue gases although
using expensive carbon sources for sulfate reduction (Gasiorek 1994; Philip & Deshusses
2003; Qian et al. 2015). The feasibility of the process proposed in this work lies in the utilization
of organic waste effluents to reduce sulfate to sulfide (as vinasse, cheese whey, pig slurry or
crude glycerol) minimizing methanogenesis. Waste organic effluents have been used in many
works to produce methane in anaerobic digestors (Filidei et al. 2003; Astals et al. 2012; Baba
et al. 2013) but few studies has been focused on minimizing methane production for sulfate
reduction (Xu et al. 2012). Then, the aim of this study is to demonstrate the feasibility of the
process proposed testing maximum capabilities of each of the stages and getting valuable
data from different operating scenarios.
2.1. Granular sludge acclimation and activity tests

Competitive biological sulfate reduction requires the use of an inexpensive carbon source.
Before the start-up of the sulfate reducing reactor, screening of different C sources (obtained
as by-products or waste effluents from industrial processes) was performed. The aim of this
screening was to select the C source presenting the higher activity with lower methane
production to perform sulfate reduction. Pig slurry, crude glycerol, cheese whey and vinasses
were selected to perform activity tests in sealed bottles and decide which C source was the
most adequate for sulfate reduction. Granular sludge obtained from an upflow anaerobic
sludge (UASB) digester at a paper and pulp industry was used to perform the activity tests.
Three sealed bottles were prepared for each C source tested, which contained 5 g of granular
sludge, mineral medium and initial COD and sulfate concentrations of around 700 mg O2 L-1
and 80 mg SO42--S L-1, respectively. The composition of the mineral medium was (g L-1): NH4Cl
(0.5), K2HPO4 (3.5), 800 mL of distilled water and 200 mL of tap water, adjusted to pH=8.5
with NaOH. The total volume was 250 mL, 150 mL of liquid phase and 100 mL of headspace.
The bottles were continuously shaken at 30ºC.
Three weeks before the activity tests, biomass was acclimated to each C source by renewing
every three days the supernatant and diffusing nitrogen in the headspace of the bottles before
their sealing to ensure anaerobic conditions. The conditions set during the acclimation period
and the activity tests were the same although activity tests were performed during 24h. The
procedure followed before starting the activity tests consisted of a first wash of the acclimated
granular sludge with mineral medium free of sulfate and C source. Granular sludge was not
centrifuged but settled during the wash-up step and the supernatant decanted to avoid
52
granules damaging. Once the supernatant was clean, the granular biomass acclimated to
each C source was mixed and re-distributed in three clean serum bottles to obtain similar
biomass concentrations and comparable activities. The bottles were sealed, deoxygenated
and pulses of C source and sulfate were added to each bottle to begin the activity tests. During
activity tests the bottles were sampled to monitor sulfate, COD and methane concentrations.
Six liquid samples were taken and analyzed during the activity test from two of the bottles
prepared for each C source tested. The third bottle served to monitor methane production
through the increase of the headspace pressure (Centrepoint Electronics). Initial and final
samples from the liquid phase of the third bottle were also analyzed to ensure that similar
activities were obtained in all the bottles for comparison purposes.
2.2. Experimental setup for sulfate reduction

The schematic of the experimental setup used in this study is presented in Figure 1.
H2S,
CH4
Redox
pH
TOC
H2S, HS-
SO42-
Carbon
source
Figure 1. Experimental setup designed for sulfate reduction.
The UASB type reactor was made of glass with a reaction volume of 1L and a total volume of
2L. The reactor consisted of two different sections, i.e. a lower section (riser) containing the
sludge blanket with a diameter (D) of 0.05 m and the upper section as biomass, liquid and gas
separator (D=0.11 m). Mineral medium was pumped, together with the organic source
selected after the screening stage, from the bottom to the top of the reactor (upflow of the
medium through the sludge blanket). The composition of the mineral medium was (g L-1):
NH4Cl (0.5), K2HPO4 (3.5) and Na2SO4 (0.2) dissolved in tap water and adjusted to pH=9.0
with NaOH (2 M). A total flow rate of 0.5 L h-1 was pumped into the system (upflow velocity of
0.25 m h-1), which consisted of a mixture of the mineral medium and the C source. The flow
rate of the organic effluent was set at 1.2 mL h-1. Hence the C source was diluted to adjust the
inlet TOC concentration. T and pH in UASB were not controlled but monitored. The gas was
collected in a sample Tedlar bag of 5 L (SKC Inc.) to analyze H2S concentration, methane
53
concentration and methane flow rate. Inlet and outlet flows were also sampled every two or
three days to analyze TOC and sulfur compounds (sulfate, thiosulfate and sulfide). VFA and
granules size distribution were also analyzed occasionally at three different column heights
(H=5, 25 and 44 cm) to study the sludge bed stratification.
2.3. Start-up and operating conditions of the process

The UASB was inoculated with anaerobic granular sludge from an UASB for biogas production
of a paper and pulp industry. The operation of the reactor lasted 4 months. Sulfate inlet
concentration and hydraulic residence time (HRT) were set at 220 mg S-SO42- L-1 and 2h,
respectively. The C source selected from the screening tests was fed at different organic
loading rates (OLR). Table 1 shows the conditions tested along the study.
Table 3. Conditions tested in the system during 140 days of UASB operation.
Time HRT Sulfate SLR TOC DQOt OLR TOC/S
(d) (h-1) (mg S L-1) (mg S L-1 h-1) (mg C L-1) (mg O2 L-1) (mg C L-1 h-1) (g C g-1 S)
0 – 50 176±26 548±80 90.0±13.3 0.8
50 – 90 1.95±0.2 503±85 1567±264 258±43 2.3
220±17 113±16
90 – 110 9 607±119 1891±370 311±61 2.8
110 – 140 329±29 1026±90 168±15 1.5
In addition to the reactor monitoring, the granular sludge bed stratification in the UASB was
studied and the biological processes occurring in different heights of the reactor proposed.

Sulfate (SO42-) and thiosulfate (S2O32-) concentrations were determined by ion
chromatography with conductivity detection using a Dionex ICS-2000. The inorganic carbon
concentration was measured with an OI Analytical TOC Analyzer (Model 1020A). A SenTix®
82 probe (WTW) connected to a bench-top multimeter (Inolab® Multi 740 – WTW) was used
for routine T and pH measurements. Sulfide concentration was analyzed off-line with a sulfide
selective electrode (VWR International Eurolab, S.L). VFA were analyzed by gas
chromatography (Agilent Technologies, 7820A) equipped with a DB-FFA column and using a
flame ionization detector (FID) with helium as carrier gas. CH4 and H2S were analyzed by gas
chromatography (Hewlett Packard, HP 5890) equipped with a Porapack Q column and a
thermal conductivity detector (TCD) with helium as carrier gas. The average particle size was
measured by a laser particle size analysis system (Malvern MasterSizer Series 2600, Malvern
instruments Ltd., UK).
3.1. Screening of carbon sources

Activity tests performed after granular sludge acclimation to different C sources allowed
selecting the most appropriate organic waste for the biological process proposed in this work.
Sulfate, COD and methane concentration profiles and sulfate uptake rate profiles obtained
from activity tests performed with pig slurry, glycerol, cheese whey and vinasse are presented
in Figure 2. As can be observed in Figure 2A and 2B the deviation associated to the activity
tests performed in two different bottles with the same C source were below 9% indicating that
the tests had similar biomass concentrations, activities and therefore were comparable. The
average and deviation calculated from initial and final concentrations of sulfate and COD took
also into account the results obtained from the bottles used to monitor methane production.
The low deviation obtained for initial and final concentrations indicated that methane
54
production profile could be correlated with sulfate and COD profiles obtained from liquid phase
monitoring.
Figure 2. Activity tests perform in sealed bottles with sulfate and different carbon wastes (Pig
slurry, glycerol, cheese whey and vinasses). A and B) Degradation profiles of sulfate and COD,
respectively, C) Methane production and D) Sulfate uptake rate for each carbon source tested.
Figure 2A and 2D revealed that pig slurry was the C source that presented a lower sulfate
uptake rate (SUR) with removal efficiency (RE) of 22.0±1.0%. For this reason pig slurry was
the first C source discarded in this screening. Vinasse, cheese whey and crude glycerol
presented satisfactory results related with sulfate reduction with removal efficiency of
68.0±2.4%, 51.2±0.9% and 42.5±0.6%, respectively. However, vinasse and cheese whey
were probably limiting sulfate reduction since at the end of the test COD concentrations were
around 50 mg O2 L-1, which indicated that higher sulfate reduction percentage could be
reached at higher COD/S ratios than 10 g O2 g-1 SO42--S. Figure 2D cleared up this assumption
since the last SUR calculated for vinasse and cheese whey were much lower than that
calculated between t=6h and t=12h, indicating that a limitation appeared during the last 12h
of activity test. This was not the case of crude glycerol that showed increasing SUR along the
activity test without limitations, although with much lower SUR (1.08±0.02 mg S L -1 h-1)
compared to those obtained until t=12h with cheese whey and vinasse (3.16±0.03 mg S L-1 h-
1
and 4.03±0.33 mg S L-1 h-1, respectively). Methane profiles in Figure 2C helped to explain
such performance. As can be observed, vinasse and cheese whey had a methane production
between 1.5-fold and 2-fold that measured with crude glycerol. This fact not only caused C
source limitation but lower selectivity of carbon for sulfate reduction. The percentage of C
source used for methane production in the case of vinasse, cheese whey and pig slurry was
70.8±4.3%, 59.1±4.1% and 60.2±1.9%, respectively, while for glycerol a 49.7±1.7% was
55
obtained. These results allowed selecting crude glycerol as the most adequate C source for
sulfate reduction since this biodiesel industry by-product, even presenting lower SUR,
produced less methane and had as a result higher efficiency of carbon utilization for sulfate
reduction.
3.2. UASB performance as a function of the C/S ratio supplied

As mentioned above, crude glycerol was used as the C source for sulfate reduction. Sulfate
reduction in the UASB was started-up with a TOC/S ratio of 0.8 g C g-1 SO42--S and then 3
different TOC/S ratios were tested while sulfate concentration was set at 220 mg SO 42--S L-1
as presented in Table 1. In Figure 3 the profiles obtained from the monitoring of some of the
species involved in the process are presented.
Figure 3A shows that after 15 days of operation the sulfate RE was 42.1%, which
corresponded to an elimination capacity of 55.8 mg SO42--S L-1 h-1. From this point until day
50, the average RE was 34.3±3.5% although sulfate reduced was not completely recovered
as sulfide. This result can be explained with the formation of elemental sulfur in the upper part
of the reactor. Elemental sulfur was produced due to the undesired diffusion of air through
silicon tubing that lead to the biological partial oxidation of sulfide under microaerophilic
conditions. This hypothesis was plausible because thiosulfate was not detected in the UASB.
This probably means that sulfite was not produced during sulfate reduction since, according
to Mora et al. (2016), thiosulfate together with elemental sulfur could be produced from the
chemical reaction between sulfite and elemental sulfur in the presence of sulfide and under
aerobic conditions with polysulfide as the intermediate of the reaction. Regarding TOC profiles
obtained during this first period of 50 days, the RE was 76.9±7.0%, which probably
corresponded to the most biodegradable fraction of crude glycerol. As reported by Hansen et
al. (2009) and Hu et al. (2012) crude glycerol can contain more than a 50% of organic matter
different from glycerol and methanol as fatty acid methyl esters, free or long chain fatty acids
and glycerides, which are complex compounds difficult to be biodegraded and even inhibitory
(Chen et al. 2008). In Table 2 the summary of RE resulting from the remaining operation
periods is presented. It must be pointed out that the second half of the sludge blanket lost the
granulation probably due to the development of sulfate reducing bacteria, the presence of
sulfide and the low upflow velocity (Liu & Tay 2004; Pol et al. 2004).
During the second period (days 50 to 90) TOC inlet concentration was increased (Table 1) to
reach a TOC/S ratio of 2.3 g C g-1 SO42--S since the previous ratio tested (0.8 g C g-1 SO42--S)
was not high enough to completely reduce sulfate to sulfide. The OLR during this period
(258±43 mg C L-1 h-1) allowed obtaining a S-RE and C-RE of 83.0±12.8% and 80.8±5.2%,
respectively, which indicated that sulfate reducing bacteria were not limited by C source
availability but also that other heterotrophic bacteria (fermentative, acetogenic and
methanogenic bacteria) were growing since methane production increased until values over
100 mL h-1 (equivalent to 98.2 mg C L-1, corresponding to 20% of the inlet TOC). The instability
of the reactor was caused by operational problems. Pumping problems of mineral medium into
the reactor due to the massive growth of heterotrophic bacteria, lead to clogging of the inlet
tubes and diffuser. Also, the inefficient separation in the upper part of the reactor, where
biomass accumulated into the gas collector, did not allowed the gas produced flowing out the
reactor to a tedlar bag. Methane concentration was more stable reaching concentrations of
85.9±6.6% due to the modification of tubing materials, which minimized the entrance of air in
the headspace of the reactor. Nonetheless, these modifications did not avoid microaerophilic
conditions in the headspace of the reactor since elemental sulfur was still produced and H2S
56
content in the tedlar bag was less than 5% of inlet sulfate. It must be mentioned that CO2 was
present in a low percentage due to the pH of the effluent, which was around 7.5.
350
A Sulfate (inlet)
Sulfate, Sulfide (mg S L-1)

300 Sulfate (UASB)
Sulfide (UASB)
250
200
150
100
50
0
0 25 50 75 100 125 150
Time (d)
1000
TOC (inlet) B
TOC (UASB)
800
TOC (mg S L-1)
600
400
200
0
0 25 50 75 100 125 150
Time (d)
200 100
C F
CH4
160 CH4 80
CH4, H2S (%)
H2S
FCH4 (mL h )
-1
120 60
80 40
40 20
0 0
0 25 50 75 100 125 150
Time (d)
Figure 3. Experimental profiles obtained from the operation monitoring of the UASB during 140 days.
A) inlet sulfate concentration and S-profiles B) inlet TOC and C-profiles from the outlet of the reactor
and C) methane flow and percentage profiles of methane and hydrogen sulfide produced in the
reactor.
57
Table 2. Summary of sulfate and TOC removal efficiencies (S-RE and C-RE, respectively) obtained
from UASB operation.
Operation periods
RE
0 – 50 days 50 – 90 days 90 - 110 days 110 – 140 days
S-RE (%) 30.5±8.0 83.0±12.8 97.1±2.9 57.6±13.5
C-RE (%) 73.6±17.0 80.8±5.2 43.0±8.1 35.3±6.6
The third operation period (days 90 to 110) corresponded to a TOC/S ratio of 2.8 g C g-1 SO42-
-S). An average S-RE of 97.1±2.9% was reached (Figure 3 and Table 2). Compared to the
previous operation period, C-RE dropped drastically, which probably indicated that the high S-
RE of sulfate and the corresponding accumulation of sulfide in the sludge blanket were
affecting the activity of the abovementioned heterotrophic bacteria. The increase of C source
utilization efficiency by sulfate reducing bacteria resulted in a maximum SUR of 109.7 mg SO42-
-S L-1 h-1, which is 2.2-fold that obtained by Bertolino et al. (2014) using glycerol as the carbon
source and comparable to those obtained by Celis-García et al. (2007) using lactate and VFAs
as C sources. A maximum C elimination capacity of 133.7 mg C L-1 h-1 was obtained for
glycerol. This result indicated that a complete sulfate reduction could be reached at TOC/S
ratios lower than 1.2 g C g-1 SO42--S. From this period the analysis of VFA at three different
heights of the sludge blanket was performed to study its stratification.
From the results obtained in the previous period a TOC/S ratio of 1.5 g C g-1 SO42--S was set
during the last operation period (110 - 140 days) to avoid the excess of organic matter in the
effluent of the reactor. The decrease of inlet glycerol concentration from 607±119 mg C L-1 to
329±29 mg C L-1 caused the lost of granulation in most of the sludge blanket, rising of biomass
and subsequently the loss of 30% of sludge blanket. S-RE and C-RE decreased notably until
reaching 57.6±13.5% and 35.3±6.6%, respectively. The process failure occurring in this period
indicated certainly that fermentative, acetogenic and methanogenic bacteria were severely
affected by this abrupt modification of the operating conditions. The methane production
decrease also confirmed the loss of methanogenic activity. Other authors also reported severe
sludge flotation in anaerobic reactors fed with crude glycerol due to the presence of long chain
fatty acids (Viana et al. 2012).
3.3. Biomass granulation and biodegradation process

Operating conditions set in the reactor (upflow velocity, SLR, OLR, pH and T) caused the
stratification of the sludge blanket. To study this stratification two analyses were performed,
the particle size distribution of the granules and the VFA along the reactor height. Results from
these analyses are shown in Figure 4. As observed in Figure 4A, the particle size distribution
obtained 10 days after the start-up of the process revealed that the granulation was maintained
in the first part of the sludge blanket (5 cm) but lost in, at least, half of reactor since the riser
had a total length of 50 cm and 50% of the particles at H=25 cm had a diameter higher than
214.6 µm, which is the limit to consider biomass aggregates as granules. The situation was
different after 110 days of operation since granulation was recovered although the difference
between granules size in the upper and lower parts of the sludge blanket was really clear
(Figure 4B). The analysis of particle size distribution revealed that half of the granules (D(0.5))
measured at H=5 cm had a diameter higher than 729.3 µm. However, half of the granules
measured at H=25 cm and H=44 cm had a diameter higher than 391.7 µm. This result can be
related with biological processes occurring along the reactor height. Analysis of VFA (Figure
4A) indicated that in the first part of the sludge blanket (H=5cm) fermentative bacteria and
58
acetogenic bacteria were predominant since glycerol was converted mainly to acetic acid and
propionic acid (Viana et al. 2012). Then acetic acid was consumed while propionic acid was
still increasing indicating that sulfate reduction was preferentially carried out with acetic acid.
Methanogenesis was probably occurring mainly at the beginning of the sludge blanket since
sulfide affects severely the methanogenic activity (Chen et al. 2008).
1000 400
A 10 days B
110 days
Acetic acid
800
Propionic acid 300
Butyric acid
VFA (mg C L-1)

600
D (0.5)
200
400
100
200
0 0
0 5 25 44
Reactor height (cm)
Figure 4.Study of sludge blanket stratification. A) particle size distribution of the granules and VFA
content along the reactor height, B) image of the sludge blanket appearance after 110 days of operation.
The following reactions could explain the biological processes involved along the reactor
height (Equations 1-5) (Rajendran et al. 2014).
It is worth mentioning that hydrogen was never detected during system operation, although it
could be produced and consumed along the sludge according to the stoichiometry depicted in
Equations 1 and 2.
59
4. CONCLUSIONS
A biological process was proposed in this study to reduce sulfate with crude glycerol in an
UASB as the first step towards the recovery of sulfur from S-rich effluents. Prior to the start-
up of the process the screening of different C sources allowed concluding that sulfate
reduction could be efficiently performed with vinasses, cheese whey and crude glycerol,
although crude glycerol produced less methane. Granular anaerobic sludge from a paper
industry was used for sulfate reduction which lost the granulation after 15 days of operation
due to the conditions set and sulfide produced. Four different TOC/S influent ratios were tested
and 100% of sulfate removal efficiency was reached for TOC/S ratios above 2.3 g C g-1 SO42-
-S, when carbon availability was not limiting the process. The stratification of the sludge
blanket was studied and the analysis of VFAs along the UASB revealed that acidogenesis and
methanogenesis occurred probably in the first half part of the reactor, where sulfide was not
present. Then sulfate reduction was the main process running, probably from acetic acid
produced in the preceding portion of sludge blanket (0<H<25 cm). Overall the process was
successfully tested during 4 months an allowed the reduction of sulfate from an S-rich effluent
using waste organic effluent such as crude glycerol.
ACKNOWLEDGMENTS
Authors are members of the GENOCOV research group (2009 SGR 815) from the Department
of Chemical, Biological and Environmental Engineering at UAB (Universitat Autònoma de
Barcelona), a unit of Biochemical Engineering of Xarxa de Referència en Biotecnologia de
Catalunya (XRB), Generalitat de Catalunya. We thank Nubiola Spain (FERRO Corporation)
for the financial support of this work.
REFERENCES
Astals S., Nolla-Ardevol V. and Mata-Alvarez J. (2012). Anaerobic co-digestion of pig manure
and crude glycerol at mesophilic conditions: Biogas and digestate. Bioresource Technology
110, 63-70.
Baba Y., Tada C., Watanabe R., Fukuda Y., Chida N. and Nakai Y. (2013). Anaerobic
digestion of crude glycerol from biodiesel manufacturing using a large-scale pilot plant:
Methane production and application of digested sludge as fertilizer. Bioresource
Technology 140, 342-8.
Celis-Garcia L. B., Razo-Flores E. and Monroy O. (2007). Performance of a down-flow
fluidized-bed reactor under sulfate reduction conditions using volatile fatty acids as electron
donors. Biotechnology and Bioengineering 97(4), 771-9.
Chen Y., Cheng J. J. and Creamer K. S. (2008). Inhibition of anaerobic digestion process: A
review. Bioresource Technology 99(10), 4044-64.
Filidei S., Masciandaro G. and Ceccanti B. (2003). Anaerobic digestion of olive oil mill
effluents: Evaluation of wastewater organic load and phytotoxicity reduction. Water Air and
Soil Pollution 145(1), 79-94.
Gasiorek J. (1994). Microbial Removal of Sulfur-Dioxide from a Gas-Stream. Fuel Processing
Technology 40(2-3), 129-38.
Gomez D. K. V., Papirio S., Kramer H. J. M. and Lens P. N. L. (2010). Metal removal
mechanisms in sulfate reducing reactors: effect of sulfide concentration. Journal of
Biotechnology 150, S263-S4.
Hansen C. F., Hernandez A., Mullan B. P., Moore K., Trezona-Murray M., King R. H. and
Pluske J. R. (2009). A chemical analysis of samples of crude glycerol from the production
of biodiesel in Australia, and the effects of feeding crude glycerol to growing-finishing pigs
60
on performance, plasma metabolites and meat quality at slaughter. Animal Production

Science 49(2), 154-61.
Klimont Z., Smith S. J. and Cofala J. (2013). The last decade of global anthropogenic sulfur
dioxide: 2000-2011 emissions. Environmental Research Letters 8(1).
Liu Y. and Tay J. H. (2004). State of the art of biogranulation technology for wastewater
treatment. Biotechnology Advances 22(7), 533-63.
Papirio S., Villa-Gomez D. K., Esposito G., Pirozzi F. and Lens P. N. L. (2013). Acid Mine
Drainage Treatment in Fluidized-Bed Bioreactors by Sulfate-Reducing Bacteria: A Critical
Review. Critical Reviews in Environmental Science and Technology 43(23), 2545-80.
Philip L. and Deshusses M. A. (2003). Sulfur dioxide treatment from flue gases using a
biotrickling filter - Bioreactor system. Environmental Science & Technology 37(9), 1978-82.
Pol L. W. H., Lopes S. I. D., Lettinga G. and Lens P. N. L. (2004). Anaerobic sludge
granulation. Water Research 38(6), 1376-89.
Qian J., Liu R. L., Wei L., Lu H. and Chen G. H. (2015). System evaluation and microbial
analysis of a sulfur cycle-based wastewater treatment process for Co-treatment of simple
wet flue gas desulfurization wastes with freshwater sewage. Water Research 80, 189-99.
Rajendran K., Kankanala H. R., Lundin M. and Taherzadeh M. J. (2014). A novel process
simulation model (PSM) for anaerobic digestion using Aspen Plus. Bioresource Technology
168, 7-13.
Scherson Y. D. and Criddle C. S. (2014). Recovery of Freshwater from Wastewater:
Upgrading Process Configurations To Maximize Energy Recovery and Minimize Residuals.
Environmental Science & Technology 48(15), 8420-32.
Srivastava R. K. and Jozewicz W. (2001). Flue gas desulfurization: The state of the art. Journal
of the Air & Waste Management Association 51(12), 1676-88.
Viana M. B., Freitas A. V., Leitão R. C., Pinto G. A. S. and Santaella S. T. (2012). Anaerobic
digestion of crude glycerol: a review. Environmental Technology Reviews 1(1), 81-92.
Villa-Gomez D. K., Papirio S., van Hullebusch E. D., Farges F., Nikitenko S., Kramer H. and
Lens P. N. L. (2012). Influence of sulfide concentration and macronutrients on the
characteristics of metal precipitates relevant to metal recovery in bioreactors. Bioresource
Technology 110, 26-34.
Xu X. J., Chen C., Wang A. J., Fang N., Yuan Y., Ren N. Q. and Lee D. J. (2012). Enhanced
elementary sulfur recovery in integrated sulfate-reducing, sulfur-producing rector under
micro-aerobic condition. Bioresource Technology 116, 517-21.
Yuan H. P. and Zhu N. W. (2016). Progress in inhibition mechanisms and process control of
intermediates and by-products in sewage sludge anaerobic digestion. Renewable &
Sustainable Energy Reviews 58, 429-38.
61
Fluidised Bed Gasification of Sewage Sludge using

different Bed Materials
Virginia Pérez López1*, José María Murillo Laplaza1, Alfonso Pascual Delgado1, José
María Sánchez-Hervás2 and Raquel Ramos Casado1
1CEDER-CIEMAT, Autovía A15, salida 56. 42290 Lubia (Soria), Spain.
2CIEMAT, Avenida Complutense, 40. 28040 Madrid, Spain.
* Corresponding author: virginia.perez@ciemat.es, phone: +34975281013
ABSTRACT
The aim of this work is to study the gas quality, the tar production, the heating value of the
gas, the gas yield, the carbon conversion and the efficiency of the sewage sludge gasification
process using three different bed materials: silica, olivine and activated olivine. Air gasification
tests of digested dry sewage sludge from a waste-water treatment plant have been performed
in a pilot bubbling fluidized bed gasifier under atmospheric pressure and in auto-thermal
operation, without the use of any external heating. Gasification tests have been conducted in
a temperature range of 785-810 ºC, using different bed materials. Air gasification of sewage
sludge produces a low heating value gas with high tar content. A better quality gas has been
produced using olivine as bed material (activated or not) in comparison with the silica use.
Keywords: Gasification, Bed, Sludge, Tar.
1. INTRODUCTION
The management of sewage sludge is a big problem all over the world because of its large
quantities and harmful impact on the environment [1]. The use of sewage sludge in agricultural
lands represents 80 % of all sewage sludge utilized in Spain [2]. Nevertheless, over the years
the continuity of this practice has resulted in an accumulation of heavy metals in the soil. In
addition, landfilling represents around 20 % of the sewage sludge produced in Europe.
According to European regulations, management methods involving dumping are now being
replaced by methods leading to waste stabilization and safe recycling [3, 4]. Gasification is a
promising sewage sludge-to-energy method considering the increasing price of land for
landfills [4]. Gasification is a thermal process by which the sewage sludge is converted to a
combustible gas producing ash, under a reducing atmosphere. This technology recovers
energy from the sewage sludge by producing a synthesis gas that can be used to obtain
thermal energy, electricity, chemical products, fertilisers and fuels. [5]. The syngas is a mixture
of H2, CO, CO2, CH4 and gaseous hydrocarbons. When air is used as gasifying agent, the
syngas contains N2. Raw gas produced from gasification also contains contaminants that must
be reduced to meet the process requirements and the pollution control regulations. Tars are
condensable hydrocarbons with a potential to foul filters, lines and engines, as well as to
deactivate catalysts. For those reasons, tar is a universal challenge of gasification [6].
Tar reduction can be achieved by different methods which can be categorized in two types,
depending on the location where tar is removed: Primary methods, taking place in the gasifier
itself; and secondary methods, taking place outside the gasifier [7]. The main primary
measures are: a proper selection of the operating conditions; use of a proper bed additive or
catalyst; a proper gasifier design [6]; and the use of different gasifying agents. The use of
certain bed materials with catalytic properties in the gasifier increases the speed of the
62
hydrocarbon reforming reactions [10]. Dolomite, olivine, alumina, alkali catalysts, nickel
catalyst are effective for producing hydrogen and eliminating tars under fluidisation conditions
[10-12]. Olivine is a natural mineral whose catalytic activity is mainly due to the presence of
magnesium and iron oxides [13]. It is consider a good mineral to be used in BFB as bed
material because of its high mechanical strength [14]. Olivine catalytic properties can be
enhanced with thermal pre-treatment of the material. This activation of the olivine increases
the Fe content and Fe-phase at the particle surface [15]. Different activation temperatures
have been studied from 900 to 1600 ºC. Several studies have been performed using a wide
range of activation times from 3 to 20 hours. Inert, reduction and oxidant atmosphere can be
used for the activation process of the olivine [13, 15, 16].

Digested Dry Sewage Sludge (DDSS) from a waste water treatment plant has been used for
the gasification tests. Its chemical characterization is shown in Table 4.
Table 4. Chemical characterization of the DDSS used in the gasification tests.

Parameter Unit Value
Moisture % w.b. 9.4
Proximate analysis
Ash % d.b. 41.2
Volatile % d.b. 53.5
Fixed carbon % d.b. 5.3
Ultimate analysis
C % d.b. 31.5
H % d.b. 4.7
N % d.b. 4.8
S % d.b. 1.0
Cl % d.b. 0.06
O % d.b. 16.8
Heating value
HHV MJ/kg d.b. 14.1
LHV MJ/kg d.b. 13.1
An experimental set-up located at the Centre of Development of Renewable Energy Sources

(CEDER), belonging to the Research Centre for Energy, Environment and Technology
(CIEMAT), has been used to study the gas composition of sewage sludge gasification using
different bed materials. The plant is shown in Figure 5. A bubbling fluidized bed gasifier (BFB)
has been operated in auto-thermal mode and at atmospheric pressure. It is 3.0 m high, with
an inner diameter of 0.3 m. It is internally coated with a layer of refractory concrete with an
outer diameter of 0.720 m. There is also an overflow pipe that is used to assure a constant
fluidized bed height. Under the distributor, a plenum reduces the velocity of the incoming air
to ensure a uniform distribution when entering the nozzles. The air for gasification is supplied
by a roots blower. A preheater with propane is used only to perform the start-up. During the
gasification process, the auto-thermal conditions maintain the temperature around 800 ºC,
without the necessity of external heating sources.
The feeding system consists of two hoopers and two helical screws, one between the two
hoopers and the other to feed the fuel into the reactor. During the test carried out for this work,
a capacity of 35 kg/h of DDSS has been used. Another screw feeder is used to refill the silica
63
sand as bed material during the tests and a second one is used to remove the mixture of ash
and bed material from the reactor.
flare
Bed material Gas
screw feeder
Cyclone
Gas analyser Tar sampling
Gasifier Ash cyclone

Sewage Sludge bin
Air heater ash bed

air
Air blower propane

Ash bed screw Ash bed
air discharge bin
Figure 5. Experimental set-up
A Supervisory Control and Data Acquisition (SCADA) system is used to operate the plant. The
produced gas leaves the reactor by the top of the freeboard and passes through a high
efficiency cyclone, where most of the particles are collected. A gas pipeline connects the
cyclone to the flare. The system includes two gas sampling points located after the gasifier,
with the aim of evaluating the gas composition and the tar content. The gas analyser used
includes a NDIR to measure CO, CO2 and CH4, a thermal conductivity analyser to determine
H2 and a paramagnetic analyser for O2. Tars are also determined following the Technical
Specification CEN/TS 15439:2006 [17]. Tar compounds can be classified in 5 different classes
according to the criteria of solubility and condensation indicated by the Energy Research
Centre of the Netherlands (ECN) [18, 19]:
- Class-1: GC undetectable tars. This class includes the heaviest tars condensing at
high temperatures even at very low concentrations.
- Class-2: Heterocyclic components (like phenol, pyridine, cresol). These are
components that generally exhibit high water solubility, due to their polarity.
- Class-3: Aromatic components. Light hydrocarbons that are not important regarding
their condensation and water solubility issues.
- Class-4: Light poly-aromatic hydrocarbons (2-3 rings PAH’s). These components
condense at relatively high concentrations and intermediate temperatures.
- Class-5: Heavy poly-aromatic hydrocarbons (4-5 rings PAH’s). These components
condense at relatively high temperature and low concentrations.
64
It should be mentioned that the gravimetric method used to determinate Class-1 tars can also
be used to determinate compounds heavier than phenantrene. The lightest compounds are
lost in the distillation process used to estimate gravimetric tars [19]. Nevertheless, Class-1 tars
were not estimated in this work. Moreover, benzene is one of the components that is
generated in the highest proportion and it is included in Class-3 tars. The gas dew point was
calculated using the complete model developed and validated by the ECN [18].
This work has covered air gasification of sewage sludge using three different bed materials
(silica, olivine and activated olivine) in order to study its influence in the gasification process.
Gasification tests have been carried out with air as gasifying agent in auto-thermal conditions
and at atmospheric pressure. The study has been performed at around 800 ºC and using an
equivalence ratio (ER) of 0.26-0.35. The equivalence ratio is defined as the quotient between
the airflow rate divided by the fuel flow rate in real conditions and stoichiometric conditions.
Silica, olivine and activated olivine have been used as bed materials to study the influence of
the different bed materials. The use of olivine and activated olivine as bed material has
catalytic properties in tar reduction comparing to using an inert material such as silica sand.
Activated olivine has been obtained at 900 ºC during 5 hours in a reducing atmosphere. Gas
quality has been determined after the gasifier, using the gas analyser and following the
Technical Specification for sampling and analysis of tar [17]. The gas yield, the carbon
conversion and the efficiency of the gasification process has been also studied.

The gasification test conditions, the gas composition and the tar content can be seen in Table
5. As can be seen, the operation has been carried out at ER values between 0.26 and 0.35,
in order to maintain the temperature around 800 ºC.
Table 5. Operating conditions and results of the gasification tests

Activated
Parameter Unit Silica Olivine
Olivine
Operating conditions
Bed Temperature ºC 786 793 807
ER - 0.26 0.29 0.35
Gas composition
H2 % b.s. 5.65 5.70 6.32
CO % b.s. 5.97 6.81 6.92
CO2 % b.s. 14.62 14.53 15.16
CH4 % b.s. 2.57 2.07 2.49
LHVgas MJ/Nm3 2.45 2.37 2.62
Tar content
Total tar g/kg fuel d.a.f. 28.35 12.51 10.71
Total tar g/Nm3 8.97 3.78 2.99
3
Class-2 tar g/Nm 1.09 0.17 0.44
Class-3 tar g/Nm3 6.66 2.80 1.76
Class-4 tar g/Nm3 1.17 0.76 0.77
Class-5 tar g/Nm3 0.06 0.04 0.02
Tar dew point ºC 134 126.5 126
Efficiency
Ygas Nm3/kg fuel d.a.f. 3.16 3.31 3.58
CGE % 35.6 36.0 43.0
Xc % 66.3 70.2 79.7
65
Figure 6 shows the gas composition during the gasification process using different bed
materials. The use of olivine increases the CO content and decreases the CH4 content
comparing it to the use of silica. The use of activated olivine increases the H2, CO and CO2
contents and maintains the CH4 content comparing it to the use of silica or natural olivine.
Figure 6. Gas composition of the gas produced in sewage sludge gasification
Figure 7 shows the lower heating value (LHV) of the gas obtained. The LHV has been
calculated considering the H2, CO and CH4 contents. On one hand, the use of natural olivine
decreases the LHV in the syngas by 3 % comparing it to the use of silica. On the other hand,
it can be seen how the use of activated olivine increases the LHV in the syngas by 7 %
comparing it to the use of silica.
Figure 7. Low Heating value of the gas produced in sewage sludge gasification
Not only gas composition changes with the use of different bed materials, but also a variation
in tar content is noticed. The effect of using catalytic bed materials can be seen in Figure 8.
The use of natural olivine and the use of activated olivine reduce the tar content in a 58 and a
67 %, respectively, regarding the results founded when using silica sand.
66
Figure 8. Tar content in the gas produced in sewage sludge gasification
Figure 9 shows the tar content results for each one of the classes of tar. As can be seen in
the figure, the contents of all classes of tars obtained with natural olivine and activated olivine
are lower than with silica. The maximum reduction percentages are found in Class-2 tars,
reaching 84% with olivine and 60% with activated olivine. In Class-3 tars, they have reached
58 % with olivine and 74 % with activated olivine. Class-4 tars are reduced 34-35% with the
use of activated olivine and olivine, respectively. Class-5 tars are also reduced 33% with the
use of natural olivine and 67% with the use of activated olivine.
Figure 9. Class-2, Class-3, Class-4 and Class-5 tar contents in the gas produced in sewage sludge
gasification
The tar dew point of the gas changes with the use of different bed materials, as a consequence
of the amount of the different tar classes (Figure 10). The use of natural olivine and activated
olivine has a clear effect on lowering the tar dew point, and even more for activated olivine,
which, as previously stated, shows lower values of Class-5 tars.
67
Figure 10. Tar dew point in the gas produced in sewage sludge gasification
The gas yield (Ygas) of the gasification process is defined as the gas flow produced divided by
the sludge flow consumed. Figure 11 compares the gas yield of the gasification tests with
silica, natural olivine and activated olivine. The maximum Ygas values were found with activated
olivine.
Figure 11. Gas yield of the sewage sludge gasification process
The cold gas efficiency (CGE) of the gasification process is defined as the combustion heating
of the gas produced divided by the combustion heating of the sludge consumed. Figure 12
compares the cold gas efficiency obtained in the tests with silica, natural olivine and activated
olivine. The maximum CGE values were found with activated olivine.
68
Figure 12. Cold Gas Efficiency of the sewage sludge gasification process
The carbon conversion (Xc) of the gasification process is defined as the weight of carbon in
the produced gas divided by the weight of carbon in the sludge introduced in the gasifier. The
variation in the carbon conversion is shown in Figure 13. The use of natural olivine and
activated olivine increase the carbon conversion.
Figure 13. Carbon conversion of the sewage sludge gasification process
4. CONCLUSIONS
 This study sets out the results obtained from conducting sewage sludge gasification
tests on an auto-thermal BFB reactor at pilot scale. The effect of using different bed
materials has been studied.
 The use of olivine increases the CO content and decreases the CH4 content and the
LHV comparing it to the use of silica. On the contrary, the use of activated olivine
enhances the gas composition increasing H2 and CO and its LHV.
 Regarding the tar content the use of natural or activated olivine reduces the total tar
content and all the classes of tar. As a result of the tar reduction, the tar dew point is
also reduced.
 In addition, the efficiency of the gasification process is better when natural or activated
olivine is used as bed material.
 As a final remark, the authors considers that it is important to emphasise that this
results need to be completed doing replications under different gasification conditions.
69
ACKNOWLEDGMENTS
The authors acknowledge the Ministry of Science and Innovation of the Spanish Government
for the financial support and CADAGUA Company, partner in the R&D project about sewage
sludge gasification named “Application of advanced technologies for energetic and process
optimization of the treatment of a Station Wastewater Treatment”.
NOMENCLATURE
ASTM American Society for Testing and Materials

BFB Bubbling Fluidized Bed
CEDER Centre of Development of Renewable Energy Sources
CEN European Committee for Standardization
CGE cold gas efficiency
CIEMAT Research Centre for Energy, Environment and Technology
d.a.f. dry ash free basis
d.b. dry basis
DDSS Digested Dry Sewage Sludge
ECN Energy Research Centre of the Netherlands
ER Equivalence ratio
HHV High Heating Value
LHV Low Heating Value
n.d. non determined
NDIR Nondispersive Infrared Sensor
PAH Hydrocarbon Poly-aromatic
SCADA Supervisory Control and Data Acquisition
TS Technical Specification
w.b. wet basis
Xc carbon conversion
Ygas gas yield
REFERENCES
[1] Kliopova I, K Makarskienė (2015) Improving material and energy recovery from the sewage
sludge and biomass residues. Waste Management 36:269-276.
[2] Ministerio de Agricultura, Alimentación y Medioambiente (2016) MAGRAMA.
http://www.magrama.es.
[3] Cieślik BM, Namieśnik J, Konieczka P (2015) Review of sewage sludge management:
standards, regulations and analytical methods. Journal of Cleaner Production 90:1-15.
[4] Samolada MC, Zabaniotou AA (2014) Comparative assessment of municipal sewage
sludge incineration, gasification and pyrolysis for a sustainable sludge-to-energy
management in Greece. Waste Management 34(2):411-420.
[5] Manara P, Zabaniotou A (2012) Towards sewage sludge based biofuels via
thermochemical conversion – A review. Renewable and Sustainable Energy Reviews
16(5):2566-2582.
[6] Woolcock PJ, Brown RC (2013) A review of cleaning technologies for biomass-derived
syngas. Biomass and Bioenergy 52(0):54-84.
[7] Devi L, Ptasinski KJ, Janssen FJJG (2003) A review of the primary measures for tar
elimination. Biomass and Bioenergy 24(2):125-140.
[8] Roche E, de Andrés JM, Narros A, Rodríguez ME (2014) Air and air-steam gasification of
sewage sludge. The influence of dolomite and throughput in tar production and
composition. Fuel 115(0): 54-61.
[9] Campoy M, Gómez-Barea A. Ollero P, Nilsson S (2014) Gasification of wastes in a pilot
fluidized bed gasifier. Fuel Processing Technology 121(0):63-69.
[10] Sutton D, Kelleher B, Ross JRH (2001) Review of literature on catalysts for biomass
gasification. Fuel Processing Technology 73(3): 155-173.
70
[11] de Andrés JM, Narros A, Rodríguez ME (2011) Behaviour of dolomite, olivine and alumina
as primary catalysts in air–steam gasification of sewage sludge. Fuel 90(2):521-527.
[12] Koppatz S, Pfeifer C, Hofbauer H (2011) Comparison of the performance behaviour of
silica sand and olivine in a dual fluidised bed reactor system for steam gasification of
biomass at pilot plant scale. Chemical Engineering Journal 175(0):468-483.
[13] Shen Y, Yoshikawa K (2013) Recent progresses in catalytic tar elimination during biomass
gasification or pyrolysis—A review. Renewable and Sustainable Energy Reviews
21(0):371-392.
[14] Corella J, Toledo JM, Padilla R (2004) Olivine or Dolomite as In-Bed Additive in Biomass
Gasification with Air in a Fluidized Bed: Which Is Better? Energy & Fuels 18(3):713-720.
[15] Devi L, Craje M, Thüne P, Ptasinski K Janssen FJJG (2005) Olivine as tar removal catalyst
for biomass gasifiers: Catalyst characterization. Applied Catalysis A: General, 294(1):
68-79.
[16] Christodoulou C, Grimekis D, Panopoulus DD, Pachatouridou EP, Iliopoulou EF, Kakaras
E (2014) Comparing calcined and un-treated olivine as bed materials for tar reduction in
fluidized bed gasification. Fuel Processing Technology 124(0):275-285.
[17] CEN/TS 15439:2006. Biomass gasification -Tar and particles in product gases - Sampling
and analysis.
[18] Energy Research Centre of the Netherlands (2016) Thersites: The ECN tar dew point site.
http://www.thersites.nl/.
[19] Van Paasen SVB, Kiel JHA (2004) Tar formation in a fluidised-bed gasifier. Impact of fuel
properties and operating conditions. ECN-C-04-013.
71
The role of nitrogen availability on methane

abatement biotechnologies
Juan Carlos Lópeza, Raúl Muñoza, Guillermo Quijanoa,b*
a Department of Chemical Engineering and Environmental Technology, University of Valladolid, Dr.

Mergelina, s/n, 47011, Valladolid, Spain. Tel. +34 983186424.
b Agrarian Engineering School, University of Valladolid - Soria Campus, C/Universidad, s/n, 42004,
Soria, Spain. Tel. +34 975129404.

*Corresponding author: gquijano@iq.uva.es, Tel. +34 975129404, Fax: +34 975129401.
ABSTRACT
Despite its short atmospheric residence time of 10 years, CH4 is considered nowadays a
greenhouse gas (GHG) more harmful than CO2, as it presents a 20-30 times higher
effectiveness to absorb long-wave radiation and it is more active in tropospheric ozone
production. Natural and anthropogenic CH4 sources (347 and 331 Tg y-1, respectively)
contribute equitably to the overall atmospheric CH4 releases, 55% of anthropogenic emissions
(derived from old landfills, coal mines, livestock farms) being below the lower explosion limit
for CH4 (< 5% v/v). Thus, biotechnologies have arisen as a promising tool to mitigate diluted
CH4 emissions, since they entail lower operating costs and environmental impacts than
physical-chemical treatment technologies. However, besides considering mass transfer
phenomena, biological performance must be considered during the implementation of CH4
abatement biotechnologies. In this context, the role of N availability on the activity of methane-
oxidizing bacteria (MOBs) is important to optimize the operation of bioreactors treating CH 4.
The few studies addressing the effect of N concentration on CH4 abatement performance were
performed in biofilm-based systems and, consequently, diffusional events might lead to
misunderstandings on the effect of N on CH4 biodegradation. To avoid this limitation, in this
work the effect of N concentration (5-70 gN m-3) on the CH4 abatement performance was
evaluated in a stirred tank reactor. It was observed that the lowest N concentration assayed
(5 g m-3) supported the highest specific CH4 elimination and CO2 production rates (0.046 and
0.028 g gbiomass-1 h-1, respectively), while it supported the lowest biomass production and
specific N consumption rates.
Keywords: Biotechnology, GHG, Methane, Mitigation, Nitrogen.
1. INTRODUCTION
Despite its relatively short atmospheric residence time of 10 years, CH4 is considered
nowadays a greenhouse gas (GHG) more harmful than CO2, as it presents a 20-30 times
higher effectiveness to absorb long-wave radiation and it is more active in tropospheric ozone
production [1,2]. Natural and anthropogenic CH4 sources (347 and 331 Tg year-1, respectively)
contribute equitably to the overall atmospheric CH4 releases, 55% of anthropogenic emissions
(derived from old landfills, coal mines and livestock farms) being below the lower explosion
limit for CH4 (< 5% v/v) [3, 4]. Nowadays, flaring constitutes the simplest and most extended
way to avoid direct CH4 releases to the atmosphere [5]. However, the treatment of such diluted
CH4 emissions by flaring requires the use of a support fuel, thus decreasing the cost-
efectiveness and sustainability of the process.
CH4 abatement biotechnologies, based on the activity of methane-oxidizing bacteria (MOBs),
represent an effective and sustainable alternative to treat diluted CH4 emissions [6,7,8]. Due
to the low aqueous solubility of CH4 (dimensionless Henry’s law constant of 30 at 25ºC and 1
atm), the CH4 mass transfer from the gas to the liquid phase is commonly the main process
governing the CH4 oxidation performance in biotechnologies [6,9]. However, nutrient
limitations must be also avoided to ensure a correct performance of the MOB community [9].
In this regard, the role of nitrogen on MOBs has been extensively discussed in the literature
72
as it is considered to play a key role on the optimal operation of biological reactors treating
CH4 [10-14]. It has been estimated that MOBs have a relatively high nitrogen requirement of
0.25 mole of nitrogen per mole of assimilated C-CH4 [15]. Moreover, previous studies on CH4
abatement carried out in biofilm bioreactors such as biofilters (BFs) and biotrickling filters
(BTFs) reported significantly different values for the optimum N-NO3- concentration in the
irrigation liquid, which ranged from 500 to 750 g m -3 [10, 11]. It must be stressed that these
important differences in the optimum N-NO3- concentration so far reported in biofilm
bioreactors might be due to both the CH4 inlet load applied and the nitrogen diffusion
phenomena occurring in these systems, the later being mainly influenced by the thickness of
the biofilm and the hydrodynamic conditions [16]. Therefore, it is not surprising to find large
discrepancies in terms of optimum N-NO3- concentration in biofilm bioreactors since
comparative assessments under identical biofilm thickness and hydrodynamic conditions are
difficult to perform. Thus, a fundamental understanding on the effect of N-NO3- on CH4
abatement biotechnologies must be evaluated in well-mixed bioreactors with suspended
biomass rather than in the configurations mentioned above.
In this study, the effect of N-NO3- concentration on the CH4 abatement performance of a
methanotrophic community was assessed in a stirred tank reactor (STR). The limiting step of
the biodegradation process was determined, while the CH4 abatement performance was
evaluated through the specific elimination capacity (EC), CO2 and biomass production (PCO2
and PX, respectively), and the N-NO3- consumption rate (rN-NO3-).

2.1 Chemicals and inoculum preparation
The mineral salt medium (MSM; final pH, 7.0) used during the trials was a modified Brunner
medium consisting of (g L-1): Na2HPO4·12H2O 2.44, KH2PO4 1.52, MgSO4·7H2O 1.00 and
CaCl2·2H2O 0.05, along with 10 mL SL4 trace element solution (g L-1: EDTA 0.5, FeSO4·7H2O
0.2, ZnSO4·7H2O 0.01, MnCl2·4H2O 0.003, H3BO3 0.03, CoCl2 0.011, CuCl2·2H2O 0.01,
NiCl2·6H2O 0.002, Na2MoO4·2H2O 0.003) per litre of medium. Additionally, 0.6 mL L−1 of a 1
gCu2+ L−1 stock solution were added to the MSM to adjust initial Cu2+ concentration at 10 µM.
A NaNO3 stock solution of 50 g L-1 was prepared to continuously supplement the MSM with
N-NO3- to the target concentrations as well as to maintain them during the trials. All chemicals
were purchased from Panreac® (Barcelona, Spain) with a purity of at least 99%. Methane was
purchased from Abelló Linde S.A. (Barcelona, Spain) with a purity of ≥ 99.5%.
The STR was inoculated with 2% (v/v) of a mixed methanotrophic consortium mainly
composed by the MOBs Methylomicrobium agile, Methylobacter sp., Methylosarcina lacus
and Methylocystis parvus [17].
2.2 Experimental set-up and operation mode

A jacketed 1.4-L STR (Afora S.A., Spain) equipped with a Rushton turbine was used for
continuous CH4 abatement along the different trials. Agitation and temperature were
maintained at 300 rpm and 25ºC (via thermostatic bath). The MSM renewal rate was set at 50
mL d-1 (dilution rate, D = 0.05 d-1) for a liquid working volume of 1 L. The pH of the culture
broth was maintained at 7.4 ± 0.4 by periodic addition of H2SO4 (98% v/v). A Fusion 100
syringe pump (Chemyx Inc., USA) was used to continuously supply N-NO3- (from the NaNO3
stock solution) at the top of the bioreactor. The synthetic CH4-laden air emission (33.8 ± 3.4
gCH4 m-3, 4.2 ± 0.5% CH4) was obtained by mixing a pure CH4 stream with an air stream via
Watson Marlow 101U/R peristaltic pump and an EVO-10 compressor (Electro A.D. S.L.,
Spain), respectively (Fig. 1). This gas CH4-laden air was fed at 100 mL min-1 to the STR
through a 10 µm porous stainless steel diffuser located at the bottom of the tank, thus resulting
in a gas residence time (GRT) of 10 min, and a CH4 inlet load (IL) of 202.8 ± 20.1 g m-3 h-1.
The effect of N-NO3- concentration on CH4 abatement performance was evaluated through
four independent tests, which lasted up to 3 weeks (until steady-state conditions were
achieved) and differed in the N-NO3- concentration set in the culture broth: 4.2 ± 2.1, 12.4 ±
73
4.7, 41.4 ± 7.3 and 71.2 ± 7.9 g m-3. Additionally, a control test without N-NO3- was conducted
to confirm the requirement of N-NO3- to achieve a steady CH4 removal. The experiments were
initially operated for 3 days under abiotic conditions in order to rule out the occurrence of
abiotic CH4 degradation.
Gas outlet Air
Compressor
Gas inlet
Mixing chamber
Syringe pump Rotameter
Feeding
pump
Stirred holding
tank
CH4
Methane
cylinder
Figure 1. Schematic of the experimental set-up.
Gas samples were periodically drawn from the sampling ports located at the inlet and outlet
of the STR to monitor the CH4, CO2, O2 and N2 concentrations. Liquid samples were
periodically taken to quantify biomass concentration via absorbance measuarements (OD650)
and total suspended solids (TSS). Liquid samples were also daily drawn to determine pH, total
organic carbon (TOC) and total nitrogen (TN) concentrations.
2.3 Assessment of the limiting step of the CH4 biodegradation
In order to assess the limiting step during methane removal in the STR, a mass transfer/uptake
test was carried out at the end of the CH4 biodegradation experiment at 5 gN m-3 according to
Estrada et al. [18]. Briefly, CH4 inlet concentration was increased from 31.8 to 63.5 g m-3 (2-
fold increase in the IL CH4) for 6 h, which raised proportionally the CH4 concentration gradient
available for mass transfer. Thus, whether the CH4 mass transfer would be the limiting step,
then the EC would increase. On the contrary, whether the bioreaction would be the limitation
in CH4 abatement, the EC would remain without changes.
2.4 Analytical procedures
The CH4, CO2, O2 and N2 gas concentrations were determined by GC-TCD according to López
et al. (2014) [15]. The pH was determined using a pH-meter Basic 20 (Crison, Spain), while
the OD650 was measured in a Shimadzu UV-2550 UV/Vis spectrophotometer (Shimadzu,
Japan). TOC/TN concentrations were measured in a TOC-VCSH analyzer (Shimadzu, Japan)
coupled with a chemiluminescense detection TN module (TNM-1) (Shimadzu, Japan). The
determination of TSS concentration was performed according to Standard Methods [19].
3. RESULTS AND DISCUSSIONS

Figure 2 depicts how a 1.9-fold increase in the IL led to a 1.5-fold increase in both the EC and
PCO2, which confirmed that the system was limited by CH4 mass transfer from the gas phase
to the liquid phase, where bioreaction takes place. It must be stressed that the CH4 mass
transfer/uptake test was performed at the lowest nitrogen concentration here evaluated, which
constitutes the pronest condition to be limited by bioreaction, possibly due to a potential
74
limitation of nitrogen. Thus, CH4 mass transport was clearly identified as the limiting step of
the process and biological limitations were ruled out for all the nitrogen concentrations tested.
Previous studies on CH4 biodegradation in STR, BF and BTF configurations revealed that CH4
mass transfer was the limiting step of the process as a consequence of the low CH4 solubility
in water [18,20,21].
50 400
45
CH4EC, PCO2 (g m h )
-1 350
40
-3
300
IL (g m h )
-1
35
-3
250
30
25 200
20
150
15
0 1 2 3 4 5 6 7 8 9
Time (h)
Figure 2. Time course of the IL (▲, dotted line), the CH4 EC (●, continuous line)
and the PCO2 (□, dashed line) during the mass transfer limitation test.
Figure 3 shows that the methanotrophic enrichment exhibited minimum values of 2.382 g m-3
h-1 and 0.243×10-3g gbiomass-1 h-1 for PX and rN, respectively, at the lowest N concentration
applied in the culture broth (Fig. 3a). However, both PX and rN almost doubled up to 4.629 g
m-3 h-1 and 0.414×10-3 g gbiomass-1 h-1 as the N concentration was increased up to 10 g m-3, thus
revealing the dependence of biomass production on the N uptake by the consortia. Moreover,
both parameters remained roughly stable despite the higher N-NO3- concentrations applied
onwards, thus not showing significant differences in the range of 10 – 70 gN m-3.
rN (g gbiomass h (x10 ))
1.4
a 5.0
-3
1.2
4.5
PX (g m h )
-1
1.0
-1
4.0
-3
0.8
-1
3.5 0.6
3.0 0.4
2.5 0.2
*
2.0 0.0
0.08 0.04
b
CH4 EC (g gbiomass h )
-1 -1
PCO2 (g gbiomass h )
0.07
-1
0.06 0.03
-1
0.05
0.04 0.02
0.03
*
0.02 0.01
0.01
0.00 0.00
0 10 20 30 40 50 60 70
-3
N concentration (g m )
Figure 3. Influence of the N concentration applied in the STR on a) PX (○, continuous line) and
rN (■, dash-dotted line) and b) CH4 EC (♦, dash-dotted line) and PCO2 (Δ, continuous line).
75
On the other hand, the evolution of both the CH4 EC and PCO2 as a function of nitrogen
concentration in the STR showed a different trend (Fig. 3b). Thus, maximum CH4 EC and
PCO2 values of 0.046 and 0.028 g gbiomass-1 h-1, respectively, were obtained at the lowest
concentration. Thereafter, these values significantly decreased to 0.011 and 0.012 g gbiomass-1
h-1 for CH4 EC and PCO2 respectively, as the N concentration concomitantly increased up to
10 and 40 g m-3, respectively. It must be noticed that no significant differences were found for
both parameters with further increases in the N concentration up to 70 g m-3.
4. CONCLUSIONS
To the best of our knowledge, this is the first systematic study addressing the influence of N
concentration (5–70 gN m-3) on the parameters describing the methane abatement
performance in a methanotrophic stirred tank reactor. Interestingly, the lowest N condition
assayed (5 gN m-3) supported the highest specific CH4 elimination capacities and CO2
production rates, while it entailed the lowest biomass production and specific N consumption
rates. Thus, it is demonstrated within this work that the use of lower N concentrations in
methane abatement technologies results not only in the reduction of reagent costs, but also
in maximizing the mitigation of this GHG. Further work could focus on the production of high-
added value products (i.e. EPS, PHAs) under the different N conditions, and on assessments
under lower/higher N concentration ranges.
ACKNOWLEDGMENTS
This research was supported by the Spanish Ministry of Economy and Competitiveness (BES-
2013-063922 contract; CTQ2012-34949 project).
REFERENCES
[1] Intergovernmental Panel on Climate Change, Fifth assessment report: Climate change
2014, Synthesis Report. https://www.ipcc.ch/report/ar5/syr/, 2014 (accessed 08.11.15).
[2] P.J. Crutzen, The role of methane in atmospheric chemistry and climate, in: W.v.
Engelhardt, S. Leonhard-Marek, G. Breves, D. Giesecke (Eds.), Ruminant Physiology:
Digestion, Metabolism, Growth and Reproduction: Proceedings of the Eighth
International Symposium on Ruminant Physiology, Ferdinand Enke Verlag, Stuttgart,
1995, pp. 291–315.
[3] C. Kenelly, S. Gerrity, G. Collins, E. Clifford, Liquid phase optimisation in a horizontal flow
biolfim reactor (HFBR) technology for the removal of methane at low temperatures,
Chem. Eng. J. 242 (2014) 144–154.
[4] S. Kirschke, P. Bousquet, P. Ciais, M. Saunois, J.G. Canadell, E.J. Dlugokencky, P.
Bergamaschi, D. Bergmann, D.R. Blake, L. Bruhwiler, P. Cameron-Smith, S.
Castaldi, F. Chevallier, L. Feng, A. Fraser, M. Heimann, E.L. Hodson, S.
Houweling, B. Josse, P.J. Fraser, P.B. Krummel, J-F. Lamarque, R.L. Langenfelds,
C.L. Quéré, V. Naik, S. O'Doherty, P.I. Palmer, I. Pison, D. Plummer, B. Poulter, R.G.
Prinn, M. Rigby, B. Ringeval, M. Santini, M. Schmidt, D.T. Shindell, I.J. Simpson, R.
Spahni, L.P. Steele, S.A. Strode, K. Sudo, S. Szopa, G.R. van der Werf, A.
Voulgarakis, M. van Weele, R.F. Weiss, J.E. Williams, G. Zeng, Three decades of
global methane sources and sinks, Nat. Geosci. 6 (2013) 813–823.
[5] A. Ordaz, J.C. López, I. Figueroa-González, G. Quijano, Assessment of methane
biodegradation kinetics in two-phase partitioning bioreactors by pulse respirometry,
Water Res. 67 (2014) 46–54.
[6] J.C. López, G. Quijano, T.S.O Souza, J.M. Estrada, R. Lebrero, R. Muñoz, Biotechnologies
for greenhouse gases (CH4, N2O and CO2) abatement: State of the art and challenges,
Appl. Microbiol. Biotechnol. 97 (2013) 2277–2303.
76
[7] J.M. Estrada, N.J.R. Kraakman, R. Muñoz, R. Lebrero, A comparative analysis of odour
treatment technologies in wastewater treatment plants, Environ. Sci. Technol. 45 (2011)
1100–1106.
[8] M. Pawlowska, Mitigation of landfill gas emissions, CRC Press, Taylor and Francis Group,
London, UK, 2014.
[9] N.J.R. Kraakman, J. Rocha-Rios, M.C.M. van Loosdrecht, Review of mass transfer aspects
for biological gas treatment, Appl. Microbiol. Biotechnol. 91 (2011) 873–886.
[10] J. Nikiema, L. Bibeau, J. Lavoie, R. Brzezinski, J. Vigneux, M. Heitz, Biofiltration of
methane: An experimental study, Chem. Eng. J. 113 (2005) 111–117.
[11] J. Nikiema, M. Girard, R. Brzezinski, M. Heitz, Biofiltration of methane: influence of inlet
load and nitrogen concentration, Can. J. Civ. Eng 36 (2009) 1903–1910.
[12] M. Girard, A. Avalos-Ramirez, G. Buelna, M. Heitz, Biofiltration of methane at low
concentrations representative of the piggery industry – Influence of the methane and
nitrogen concentrations, Chem. Eng. J. 168 (2011) 151–158.
[13] A. Avalos-Ramirez, B.P. García-Aguilar, J.P. Jones, M. Heitz, Improvement of methane
biofiltration by the addition of non-ionic surfactants to biofilters packed with inert
materials, Process Biochem. 47 (2012) 76–82.
[14] J. Hernández, S. Gómez-Cuervo, F. Omil, EPS and SMP as stability indicators during the
biofiltration of methane emissions, Water Air Soil Poll. 226 (2015) 343.
[15] P.L.E. Bodelier, H.J. Laanbroek, Nitrogen as regulatory factor of methane oxidation in
soils and sediments. FEMS Microbiol. Ecol. 47 (2004) 265–277.
[16] P. San-Valero, S. Alcántara, J. M. Penya-Roja, F. J. Álvarez-Hornos, C. Gabaldón, A tool
for predicting the dynamic response of biotrickling filters for VOC removal, Chem. Eng.
Commun. 203 (2016) 476–487.
[17] J.C. López, G. Quijano, R. Pérez, R. Muñoz, Assessing the influence of CH4
concentration during culture enrichment on the biodegradation kinetics and population
structure. J. Environ. Manage. 146 (2014) 116–123.
[18] J.M. Estrada, R. Lebrero, G. Quijano, R. Pérez, I. Figueroa-González, P.A. García-
Encina, R. Muñoz, Methane abatement in a gas recycling biotrickling filter: Evaluating
innovative operational strategies to overcome mass transfer limitations. Chem. Eng. J.
253 (2014) 385–393.
[19] APHA, Standard methods for the examination of water and wastewater, 21st ed. American
Public Health Association, Whasington, D.C., 2005.
[20] R. Lebrero, J.C. López, I. Lehtinen, R. Pérez, G. Quijano, R. Muñoz, Exploring the
potential of fungi for methane abatement: Performance evaluation of a fungal-bacterial
biofilter, Chemosphere 144 (2016) 97–106.
[21] S. Cantera, J.M. Estrada, R. Lebrero, P. García-Encina, R. Muñoz, Comparative
performance evaluation of conventional and two-phase hydrophobic stirred tank
reactors for methane abatement: Mass transfer and biological considerations,
Biotechnol. Bioeng. 9999 (2015) 1–10.
77
Oral Presentations – Biogas

78
Assessment of Thermal Cracking Tar from Sewage

Sludge Gasification
Isabel Ortiz González1, Manuel Jesús Benito González2 and José María Sánchez
Hervás3
1 Energy Valorisation of Fuels and Waste Unit, Energy Department, CIEMAT, Av. Complutense 40,
28040 Madrid. Email: Isabel.ortiz@ciemat.es
28040 Madrid. Email: mj.benito@ciemat.es
28040 Madrid. Email: josemaria.sanchez@ciemat.es
ABSTRACT
Tar are one of the major problems in the implementation of the gasification technology
because tar easily condense, even at high temperature, blocking equipment, poison catalysts
in downstream gas upgrade processes and produce corrosion. Tar cracking and reforming
methods are attractive choice to reduce the concentration of tar in the gasification product gas
because tar can be destroyed at high temperature immediately downstream the gasifier
without generating a waste liquid stream. In this work the thermal cracking/reforming of tar will
be analysed. A simulated gasification stream with a composition 11% H2, 7% CO, 17.5% CO2,
4% CH4, 12.5% H2O and 48% N2 representative of sewage sludge air gasification was
generated. Tar model compounds concentrations fed to the reactor were 25 g·m-3 for toluene
assays, and 2 g·m-3 for naphthalene in toluene solution, whose concentrations are the
maximum usually obtained in bubbling fluidized bed gasifiers. Total conversion of toluene was
achieved at 1000 ºC while naphthalene conversion only reached 33 % at the same
temperature. The principal by-products were identified and quantified with an on-line
continuous analysis system by mass spectrometry. Although thermal conversion reduces the
amount of tar, other compounds that may be less sensitive to thermal cracking were
generated, especially with the mixture of toluene and naphthalene. Therefore, higher
temperatures, above 1100 ºC, are necessary, which may compromise economic viability of
the overall process.
Keywords: tar; thermal cracking, gasification, GC-MS.
1. INTRODUCTION
Sewage sludge is a by-product of the treatment of domestic and industrial waste water. Its
production has increased in recent years so an environmentally friendly treatment of this waste
has become an important issue (Corella et al., 2006). From this point of view, thermochemical
conversion of sewage sludge by gasification produces a low quality combustible gas while
reduces the waste volume, fixes the heavy metals and recovers part of the energy contained
in it (Manya et al., 2006; Phuphuakrat et al., 2010). However, the main drawback of this
technology is the high tar and dust content in the gasification gas. Therefore, tar reduction is
one of the most important aspects of gasification because tar can block downstream
equipment and poison catalyst (Anis et al., 2013; de Andrés et al., 2011). Tar is a very complex
heterogeneous mixture of organic compounds whose composition and amount depends on
the operation conditions of the gasification process, and is defined in the Technical
Specification CEN/TS 15439 as all organic compounds present in the gasification product gas
79
excluding gaseous hydrocarbons (C1 through C6) (CEN, 2006). Depending on the type of
gasifier used, tar content typically varies from 1 g·Nm-3 for downdraft gasifier up to 100 g·Nm-
3
for updraft gasifier; but for downstream applications, tar tolerances are very strict. For internal
combustion engines up to 100 mg·Nm-3 are allowed and less than 5 mg·Nm-3 for gas turbines
(Anis et al., 2013; Milne and Evans, 1998). Physical methods for tar removal usually require
the gasification product gas to be cooled down, leading to loss of overall thermal efficiency,
and creating a waste liquid stream that must be further treated for disposal so they are not
feasible economically for practical application (Anis et al., 2013; Min et al., 2011). Tar
reforming is a more attractive method to decrease the concentration of tar in the gasification
product gas because tar can be destroyed through reforming reactions at high temperature
immediately downstream the gasifier without creating a waste liquid stream (Min et al., 2011).
In this work, the thermal degradation of model tar compounds between 550 and 1000 ºC in a
simulated stream mimicking that obtained in air gasification of sewage sludge was evaluated.
Toluene and naphthalene were selected as tar model compounds because are good
representatives of sewage sludge gasification where the major components of tar are aromatic
compounds (Adegoroye et al., 2004; Aznar et al., 2008). Toluene was used as tar model
because is one of the most abundant (Laosiripojana et al., 2014) and it is a light aromatic
hydrocarbon with a single ring compound defined as class 3 tar by Milene and Evans (Milne
and Evans, 1998) whilst naphthalene was chosen since it is a two rings compound which is a
class 4 tar and is one of the most stable product (Anis et al., 2013).

In order to study the thermal removal of tar, a lab scale test rig was used (Figure 14). It consists
of a gas mixer unit with six mass flow controllers and one liquid mass flow controller for the
introduction of water, that generates a continuous gas stream mimicking the composition of
the gas generated in air gasification of sewage sludge in an atmospheric bubbling fluidised
bed gasifier (11% H2, 7% CO, 17.5% CO2, 4% CH4, 12.5% H2O and 48% N2).
P2
Liquid Mass
flow
H 2O R
E
Siringue Pump A
C OVEN
T
O
R
Toluene
CEM
CO CO2
N2
Mixing Unit
GC/MS
Ar CH4 H2
P1
OUT
Cold bath
Figure 14. Scheme of the experimental equipment.
80
Toluene (Merck 99.5%) in a concentration of 25 g·Nm-3 and toluene – naphthalene (Sigma-

Aldrich 99.9%) mixture (9:1) were used as tar models compounds. Two different systems for
tar feeding were used: a CEM (control evaporation mixture system) and a syringe pump. When
the experiments were carried out with toluene, the simulated gasification gas was blend with
the toluene with the CEM system to obtain the final gas stream that was feed to the reactor
unit. In the case of toluene – naphthalene mixture the syringe pump was used. The reactor
unit consists in a hot box heated at 250 ºC, an electrical furnace and a reactor of 11 mm i.d.
and 300 mm length. A SCADA system allows acquiring and controls the reaction temperature,
pressure, gas and liquid flows.
An on-line continuous analysis by mass spectrometry was used to analyse the products form
during the process. A 7890A Aglient GC with a TCD and a 5975C mass spectrometry detector
was used. This equipment has two different analysis channels. The first channel was used for
the quantification of the gasification gas and consists in three columns connected in series: a
HAYESEP Q (80/100 3ft), a HP-PLOT Q (30 m x 0.53 mm x 40 m) and a HP-PLOT 5A (30
m x 0.53 mm), connected to the TCD detector. The second channel was used to identify and
quantify the by-products formed in the thermal degradation of toluene and naphthalene. It
consists in a DB-5MS column (30 m x 0.25 mm x 0.25 m) connected to the mass
spectrometry detector.
The reactor was operated in a continuous mode. Experiments were performed increasing the
temperature of the reactor in an inert atmosphere of nitrogen. When the maximum temperature
of the experiment was reached, the simulated stream of syngas was applied and the
experiment started. The temperature of each step was maintained constant for three hours.
3.1. Thermal cracking of toluene

In the Figure 15 the toluene degradation trough the temperature in a nitrogen atmosphere and
in a simulated gasification stream are shown.
120
N2
100
Conversion (%)
80 Gasification
gas
60
40
20
0
500 600 700 800 900 1000 1100
Temperature(ºC)
Figure 15. Thermal cracking of toluene in different atmospheres.
81
When the thermal cracking process in a nitrogen atmosphere was performed, the light-off
temperature, at which 50% of conversion is reached, was close to 842 ºC, being necessary a
temperature of 1000 ºC for complete toluene conversion. Nevertheless, when the same
experiment was carried out in a simulated gasification gas stream, the light-off temperature
decreased closed to 820 ºC.
Toluene conversion was not affected significantly at temperatures above 800 ºC. However,
the conversion of toluene in a simulated gasification stream increases at low temperatures.
This fact indicates that the composition of the gasification gas improves the thermal cracking
of toluene due to a reforming contribution, promoted by the presence of steam and carbon
dioxide in the gasification stream.
Thermal degradation of toluene leads to the formation of by-products. The number of by-
products was increased when temperature was decreased.
In a nitrogen atmosphere, a total of twenty-one by-products were identified and quantified by

mass spectrometry. The main compound at 1000 ºC was benzene, due to the loss of the
methyl group of the toluene molecule, and naphthalene due to the condensation of two
aromatic rings. The temperature decrease to 900 ºC resulted in lower conversion of toluene
and rise in the number and concentration of products such as benzene, naphthalene,
phenanthrene, biphenyl, biphenylene and anthracene. At 850 ºC the increase in concentration
and number of by-products was higher, especially in the heaviest compounds such as
fluorene, phenanthrene, anthracene and fluoranthene. Below 600 ºC, the conversion of
toluene was so small, that benzene was the only compound formed.
At high temperature range, between 1000 and 850 ºC, more than 95% of by-products
concentration was ascribed to benzene, followed by naphthalene (1.3% at 1000 ºC), which
concentration decreases with the temperature. In contrast, heavier compounds such as
phenanthrene, anthracene and fluoranthene reached their maximum concentration at low
temperature (700 ºC).
Regarding to the by-products formed in the gasification atmosphere, benzene was the main
by-product formed, with a maximum concentration of 12.42 g·Nm-3 at 850 ºC. Gasification
stream composition used avoids the formation of heavy compounds so only traces of indene,
naphthalene, biphenylene, biphenyl and phenanthrene were found.
3.2. Thermal cracking of toluene and toluene-naphthalene mixture

In Figure 16, the thermal conversion of naphthalene and toluene are presented.
Toluene cracking is not altered by the presence of naphthalene, and the conversion curve is
almost the same except at temperatures below 750 ºC.
Due to the higher stability of the naphthalene molecule, its conversion was lower at all
temperature reaching only a 33% of conversion at 1000 ºC.
82
120
100
Toluene
Conversion (%) 80
Naphthalene
60
40
20
0
500 600 700 800 900 1000 1100
Temperature (ºC)
Figure 16. Thermal cracking of toluene and naphthalene mixture.
When a mixture of toluene and naphthalene was used as tar model compounds stream , the
number and concentration of by-products rises. More than 30 compounds were identified and
quantified. Benzene is also the major compound presented at temperature range studied,
dismissing naphthalene, which concentration is high, not only because its conversion was low,
but also, because it is a by-product of the degradation of toluene, much heavier compounds
like phenanthrene, anthracene, fluorene or phenylnaphthalene were found.
3.4. Inert packing influence

Two types of experiments were carried out, with an inert packing bed in the reactor, and
without it, to evaluate the possible improvement in heat transfer. The inert material chosen
was silicon carbide which has been proven as inert material for tar cracking (Anis et al., 2013;
Devi et al., 2005). The effect of the inert packing bed is shown in Figure 17. The conversion
of toluene was almost the same for both cases. For naphthalene, the conversion increased
slightly when the silicon carbide was used.
120
Toluene empty
100
Toluene CSi
Conversion (%)
80 Naphthalene
Empty
60
40
20
0
500 600 700 800 900 1000 1100
Temperature (ºC)
Figure 17. Influence of an inert packing in thermal conversion.
83
The presence of an inert packing bed diminished the amount and concentration of the by-
products found, especially the heaviest compounds such as methylanthracene,
phenylnaphthalene, fluoranthene and pyrene. This fact suggests that some reactions could
have been activated on inert surface without a clear catalytic effect.
3.3. Residence time influence

In Figure 18, the influence of the residence time is evaluated. The increase of residence time
from 5 s to 15 s (measured under normal conditions) did not produce significant differences in
toluene and naphthalene conversion.
Regarding the by-products, their concentration depends on the temperature. At high

temperature range (900 – 1000 ºC), the concentration of by-products was higher with 5 s of
residence time, but at temperatures below 850 ºC, the concentration was higher with 15 s due
to the slight increase of conversion.
In relation to the by-products generated, when residence time was decreased, the number of
by-products increased due to the contribution of secondary reactions. For this reason, at all
residence time range, the main by-product was benzene. When the residence time decreased
up to 5 s, the number and concentration of heavier compounds like phenylnaphthalene,
phenanthrene or fluoranthene was higher, while biphenyl was the major product without
considering benzene at residence time of 15 s.
120
Toluene tr 15 s
100
Toluene tr 5 s
Naphthalene tr 15 s
Conversion (%)
80
Naphthalene tr 5 s
60
40
20
0
500 600 700 800 900 1000 1100
Temperature (ºC)
Figure 18. Influence of the residence time.
4. CONCLUSIONS
Total thermal conversion of toluene was achieved at 1000 ºC, while naphthalene conversion
only reached 33 % at the same temperature. Benzene was found as the main by-product, but
some heavier compounds like phenanthrene, fluoranthene and phenylnaphthalene were
found. Principal by-products were identified and quantified with an on-line continuous analysis
system by mass spectrometry. Although thermal conversion reduces the amount of tar, other
compounds that may be less reactive to thermal cracking were generated, especially with the
mixture of toluene and naphthalene. The model sewage sludge gasification composition used
as feedstock gas, an inert packing bed of SiC and at least 15 s (normal conditions) of gas
84
residence time were established as the best conditions for toluene reforming with minimum
formation of by-products. However, naphthalene seems to be insensitive to thermal cracking
in the temperature range evaluated. Therefore, higher temperatures, above 1100 ºC, or higher
residence time are necessary which may compromise the economic viability of the process.
ACKNOWLEDGMENTS
This work is a scientific result of the project CATARSYS ENE2012-36299 entitles:
“Development of Active, Selective and Stable Tar Reforming Catalysts for Upgrading of
Biomass and Waste Gasification Streams”, financed by State secretariat for Research,
Development and Innovation (Spanish Ministry of Economy and Competitiveness).
REFERENCES
Adegoroye A., Paterson N., Li X., Morgan T., Herod A.A., Dugwell D.R., Kandiyoti R. (2004)
The characterisation of tars produced during the gasification of sewage sludge in a
spouted bed reactor. Fuel 83:1949-1960.
Anis S., Zainal Z.A., Bakar M.Z.A. (2013) Thermocatalytic treatment of biomass tar model
compounds via radio frequency. Bioresour. Technol. 136:117-125.
Aznar M., Manya J.J., García G., Sánchez J.L., Murillo M.B. (2008) Influence of freeboard
temperature, fluidization velocity, and particle size on tar production and composition
during the air gasification of sewage sludge. Energy Fuels 22:2840-2850.
CEN. (2006) Biomass gasification - Tar and particles in product gases - Sampling and
analysis, CEN.
Corella J., Toledo J.M., Molina G. (2006) Calculation of the conditions to get less than 2 g
tar/mn3 in a fluidized bed biomass gasifier. Fuel Process. Technol. 87:841-846.
de Andrés J.M., Narros A., Rodríguez M.E. (2011) Behaviour of dolomite, olivine and alumina
as primary catalysts in air-steam gasification of sewage sludge. Fuel 90:521-527.
Devi L., Ptasinski K.J., Janssen F.J.J.G. (2005) Pretreated olivine as tar removal catalyst for
biomass gasifiers: Investigation using naphthalene as model biomass tar. Fuel
Process. Technol. 86:707-730.
Laosiripojana N., Sutthisripok W., Charojrochkul S., Assabumrungrat S. (2014) Development
of Ni-Fe bimetallic based catalysts for biomass tar cracking/reforming: Effects of
catalyst support and co-fed reactants on tar conversion characteristics. Fuel Process.
Technol. 127:26-32.
Manya J.J., Sánchez J.L., Ábrego J., Gonzalo A., Arauzo J. (2006) Influence of gas residence
time and air ratio on the air gasification of dried sewage sludge in a bubbling fluidised
bed. Fuel 85:2027-2033.
Milne T.A., Evans R.J. (1998) Biomass Gasifier Tars: Their Nature, Formation, and
Conversion. NREL TP-570-25357.
Min Z., Asadullah M., Yimsiri P., Zhang S., Wu H., Li C.-Z. (2011) Catalytic reforming of tar
during gasification. Part I. Steam reforming of biomass tar using ilmenite as a catalyst.
Fuel 90:1847-1854.
Phuphuakrat T., Nipattummakul N., Namioka T., Kerdsuwan S., Yoshikawa K. (2010)
Characterization of tar content in the syngas produced in a downdraft type fixed bed
gasification system from dried sewage sludge. Fuel 89:2278-2284.
85
Feedback Control of H2S Concentration in an Anoxic

Biotrickling Filter
Javier Brito1, Fernando Almenglo1, Martín Ramírez1, José M. Gómez1, Domingo

Cantero1
javier.brito@uca.es; fernando.almenglo@uca.es; martin.ramirez@gm.uca.es;

josemanuel.montesdeoca@uca.es; domingo.cantero@gm.uca.es.
1Department of Chemical Engineering and Food Technologies. Faculty of Sciences. University of

Cadiz. 11510 Puerto Real (Cádiz), Spain. Campus de Excelencia Agroalimentario CeiA3.
ABSTRACT
Hydrogen sulfide (H2S) is the main sulfur reduced compound present in biogas. Biotrickling
filters (BTFs), aerobic and anoxic, have been showed high H2S elimination capacities.
Nevertheless, anoxic BTF have one well-known advantage; not lead to biogas dilution. On the
other hand, it requires the use of nitrate as an electron acceptor. A lab-scale anoxic BTF was
built (packed bed volume 2.8 L) and fed with make-up biogas at a flow rate of 1.1 L min–1. Inlet
H2S concentrations were from 700 to 3800 ppmv and the trickling liquid velocity (TLV) was of
10 m h–1 under parallel flow operation mode. The pH was automatically controlled at 7.4 by
NaOH (5 M) or H3PO4 (0.66 M) addition. The aim of this study was to implement a feedback
control strategy. Feedback is one of the most applied control structures, because any variable
can be set close to the set-point while correcting perturbations and changes in system,
however, even though these systems have addressed different control strategies, feedback
control studies have not been yet reported. The outlet H2S concentration was the control
variable (CV) and the nitrate flow rate was the manipulated variable (MV). Three PID tuning
methods were testing: Ziegler-Nichols, Approximate M-constrained Integral Gain Optimization
(AMIGO) and maintained oscillation. The best results were obtained using PID tuning under
maintained oscillation method with 0.72 h until full control stabilization after step response in
H2S inlet concentration and excellent control results against other H2S disturbances.
Keywords: biotrickling filter; biogas; hydrogen sulfide; feedback control
1. INTRODUCTION
Biogas is an important renewable energy source whose use helps to reduce greenhouse gas
(GHG). Nowadays, it is necessary to promote the use of renewable sources in order to ensure
a sustainable supply to the environment and meet the energy needs of today's society. Thus,
European Union in 2009/28/EC Directive, for all member States in year 2020, sets the
objective to achieve 20% and 10% of energy from renewable sources for total consumption of
energy and transport sector consumption respectively. Biogas composition depends mainly of
substrate type and digestion system, thereby concentration of mainly compounds are between
45-80% for CH4 and 20-55% for CO2. This high CH4 concentration promotes the usage of
biogas as a renewable energy source. However, the presence of some impurities such as
reduced sulfur compounds, ammonia or siloxanes reduces the biogas uses. Among those
contaminant, the hydrogen sulfide (H2S) (0.01–2%) can be considered the most important
86
and it must be removed for the energetic utilization of biogas; it is a corrosive and harmful gas
and its combustion generates SOX. In particular, biogas desulfurization can be performed by
conventional physic-chemical techniques (adsorption impregnated activated carbon, chemical
washing towers, etc.) or by biological techniques, among which include the use of biotrickling
filters (BTF) (Ramírez et al. 2015).
An efficient biogas biological desulfurization technique needs a fully adapted and developed
biomass with high degradation capacity, besides it is necessary a good contact gas-liquid
phases. Reactors for this purpose may be two phases reactors, where biomass is suspended,
or three phase in which the biomass is attached to a solid support (Ramírez et al. 2011). BTFs
for biogas desulfurization has been mainly developed in the last decade by aerobic process
(oxygen as electron acceptor) or anoxic process (nitrate and/or nitrite as electron acceptor),
both biological reactions are responsibility mainly of chemoautotrophs microorganisms (Syed
et al. 2006). A common problem for both systems (aerobic and anoxic), is elemental sulfur
accumulation in packing material because of final electron acceptor limiting conditions, it could
be reduced stimulating sulfate formation by increasing the concentration of the electron
acceptor.
The most important advantages of anoxic systems are: reduced risk of explosion, no biogas
dilution and lower mass transfer limitations for nitrate (Montebello et al. 2012; Kleerebezem et
al. 2002; Fernández et al. 2014), also it allows a higher electron acceptor concentration than
aerobic systems. However, in anoxic systems it must be considered the nature of the nitrate
used, because if a chemical reactant is used the operating costs associated and the storage
risks (strong oxidizer and contact with other material may cause a fire) must be considered.
On the other hand, the use of nitrate biologically generated from a nitrification process has the
advantage to treat two effluents (biogas with H2S and wastewater with ammonium) and
simultaneously reduces the operational costs and storage risks. The cost and availability of
large amounts of chemical nitrate reduces the applicability of the anoxic BTFs, thus is
necessary the optimization of nitrate dosage to reduce the nitrate spend and purged. For this
purpose, Almenglo et al. (2015) proposed and simulated several control strategies for an
anoxic BTF, based on the nitrate dosing mode (continuous or batch), feeding biogas (constant
or variable) and two control variables (CV) (H2S concentration in gas outlet and/or sulfide in
the recirculated liquid).
In addition, control of physical, chemical and operational parameters is very important to

achieve a high capacity of H2S oxidation to ensure proper operation of the system against
disturbances. Thus, the control of these parameters can be performed manually, programed
or automatically (Almenglo et al. 2016), among them automatic mode control is the most
desirable. Important control modes are: on-off, proportional (P); proportional-integral (PI),
proportional-integral-derivative (PID). However, it is important to emphasize that despite being
widely used PID controllers, there are not previously studies in anoxic BTF. PID process
control the most important type (>95%) (Astrom et al. 1995) and feedback controls avoid
limitation of feedforward controls which are more sensitive to unpredictable perturbations
(Kramer et al. 2007; Cervin et al. 2002). Feedback control systems are the most applied
because of is possible to maintain the process variable in a value close to the disturbance set
point (Smith et al. 1985). In feedforward control, this can be avoided since disturbance is
measured and compensated before set-point deviation, but a good process understanding is
required.
Simulation and implementation of control systems in reactors has been basically focused to
control physical variables such as temperature, pH, agitation rate, flow rates, etc. (Aguilar et
87
al 2002; Lee et al, 1991). Some application in bioreactors are control influence of temperature,
substrate concentration and pH on cell growth (Rajinikanth et al. 2010; von Meien et al. 2004)
or dissolved oxygen concentration control (Åkesson et al. 1999; Chotteau et al. 2012). In
anoxic BTFs is possible to control efficiently nitrate dossing based on oxidation reduction
potential (ORP) value (Fernández et al. 2013; Almenglo et al. 2016). An on-off control strategy
has been used in aerobic BTF to maintain oxygen concentration in exhaust gas constant
around 2% (Rodriguez et al. 2014), thus it is ensured that the biological process has the
necessary amount of oxygen and excessive dilution of biogas is avoided.
It is also possible to find examples of PID control modes applied to biogas desulfurization in
other kind of bioreactors. Ramos et al. (2014) report the use of PID feedback controller for
oxygen flow rate control using the outlet H2S concentration as control variable in an anaerobic
digester for biogas producing from sludge wastewater.
H2S outlet concentration is the variable that normally wish to control in biogas desulfurization
systems. Electrochemical type sensors are often used to get a continuous measurement of
H2S concentration (Kramer et al. 2007). Another novel method for sulfur measuring in liquid
medium has been developed by Redondo et al. (2008), this method could be implemented in
a control system allowing the sulfide concentrations measurement in both: liquid medium (0.61
to 3.200 mgS2- L-1) and gas phase (400-10000 ppmv) (Montebello et al. 2010).
The work described here concerned a study about feedback control strategies on an anoxic
BTF. The CV was the outlet H2S concentration and MV was the nitrate flow rate. Three PID
tuning methods were testing: Ziegler-Nichols rules and Approximate M-constrained Integral
Gain Optimization (AMIGO).
2.1. BTF lab scale.

A lab-scale anoxic BTF (Figure 1) was built (internal diameter: 7.14 cm) with a packed bed
volume of 2.8 L. The packing material was polypropylene pall rings 5/8” (Pall Ring Company,
UK). The BTF was fed with make-up biogas at constant flow rate of 1.1 L min–1 and empty bed
residence time (EBRT) of 117 s. The gas inlet was located at the top of the column and
therefore the BTF was running in parallel flow mode. The inlet H2S concentration was between
700-3800 ppmv (Inlet load (IL) from 28.11 to 159.6 gS m–3 h–1). The trickling liquid velocity
(TLV) was 10 m h–1 (liquid flow rate of 0.666 L min–1). Temperature was controlled by a cooling
thermostat (Lauda, Germany) using a ball Allihn Refrigerant (Alamo, Spain) in the recirculation
stream. The pH and the ORP in the recirculation stream were monitored (Multimeter 44, Crison
Instruments, Spain). The pH was controlled at 7.4 by NaOH (5 M) or H3PO4 (0.66 M) addition.
The mineral medium composition was (g L–1): NH4Cl (1), K2HPO4 (2), MgSO4·7H2O (0.8) and
trace elements. Trace element solution SL-4 (5 mL L–1) and a solution of FeSO4∙7H2O (2 g in
1 L H2SO4 0.1 N) (10 mL L–1). The composition of trace element solution SL-4 was: EDTA (0.5
g L–1), FeSO4∙7H2O (0.2 g L–1) and trace element solution SL-6 (100 mL L–1). Trace element
solution SL-6 was made of (g L–1): ZnSO4∙7H2O (0.1), MnCl2∙4H2O (0.03), H3BO3 (0.3),
CoCl2∙6H2O (0.2), CuCl2∙2H2O (0.01), NiCl2∙6H2O (0.02), Na2MoO4∙H2O (0.03) (Fernández et
al. 2013)
The mineral medium was enriched with NaNO3 (0.8-1.33 g L–1). The nitrate flow rate was the
MV and it was controlled by a feedback control using an analog peristaltic pump (PD5101
Heidolph Instruments GmbH & Co. KG, Germany).
88
Figure 19. Experimental installation setup. (1) BTF, (2) H2S sensor, (3) pH probe (4) ORP probe, (5)
Inlet gas, (6) Mass flow controller, (7) Multimeter 44, (8) PC and control system, (9) NaOH container,
(10) H3PO4 container, (11) Mineral medium+NaNO3 container, (12) Analog peristaltic pump, (13)
H3PO4 peristaltic pump, (14) NaOH peristaltic pump, (15) Recirculation pump, (16) Liquid rotameter,
(17) Thermostatic bath, (18) Liquid purge, (19) Outlet gas, (20) Gas cylinders.
2.2. Analytical methods.

Gas chromatograph (450-GC, Bruker, Spain) equipped with thermal conductivity detector
(TCD) and Poraplot Q plot FS 25mx0.53mm column was used for measure of gas
concentrations (> 1000 ppmv of H2S), at the following experimental conditions: oven
temperature: 33ºC (2 min), 33-80ºC (10ºC min–1) and injector and detector temperature:
150ºC. Gas samples at lower H2S concentration (0-100 ppmv) were measured by a gas
chromatograph (450-GC, Bruker, Spain) equipped with pulsed flame photometric detector
(PFPD) and Wcot Fused Silica 30mx0.32mm capillary column, at the following experimental
conditions: oven temperature: 35ºC (1.5 min), 33-80ºC (15ºC min–1), injector temperature
250ºC and detector temperature: 200ºC. Nevertheless, gas phase streams obtained from BTF
gas outlet were on-line monitored with an electrochemical H2S sensor (Sure-cell, Euro-Gas
Management Services, UK). The electrochemical H2S sensor has a measurement range
between 0-200 ppmv, so the outlet biogas stream was diluted with air (1.2 L min–1) before to
pass through the sensor.
In the liquid purge stream were taken samples for measurement of ionic sulfate, nitrate and
nitrite. All ionic compounds were quantified using an IC Flex Ion Chromatography (IC) system
with a Metrosep A Supp 5 - 250/4.0 column (Metrohm, Switzerland).
2.3. Tuning methods.

Three analytical methods for tuning were used, two of them based on step response: Ziegler-
Nichols rule (Ziegler et al. 1942)and Approximate M-constrained Integral Gain Optimization
(AMIGO) method (Hägglund et al. 2002). In both cases the inlet H2S concentration was 20%
increased from 1900 to 2280 ppmv (IL from 79.8 to 95.76 gS m–3 h–1). Parameters for
controllers P, PI and PID were estimated according to Hägglund et al. (2002) (Table 1).
89
Table 1. Proportional gain, integral and derivative time determination for step response method.
Ziegler-Nichols rule AMIGO
PID PI P PID PI
Proportional gain (Kp) 1.2T/(K L) 0.9T/(K L) T/(K L) 1/K(0.2+0.45T/L) 0.25T/(K L)
Integral time (I) 2L 3L - 0.4L+0.8T/(L+0.1T)L 0.8T
Derivative time (D) L/2 - - 0.5LT/(0.3L+T) -
The other one was maintained oscillation method (Ziegler et al. 1942). This method is based
on experiments executed on a feedback control under proportional control. It were tested six
proportional gain values (Kp: 0.05, 0.018, 0.015, 0.012, 0.011 and 0.01 at inlet H2S
concentration of 2280 ppmv (IL of 95.76 gS m-3 h-1).The tuning formula is based on the
empirical knowledge of the ultimate gain (Ku), and ultimate period (Tu) (Hang et al. 1991), as
shown in Table 2.
Table 2. Proportional gain, integral and derivative time determination for maintained oscillation
method.
Ziegler-Nichols rule
PID PI P
Proportional gain (Kp) 0.600 Ku 0.450 Ku 0.500 Ku
Integral time (i) 0.500 Tu 0.850 Tu -
Derivative time (d) 0.125 Tu - -
So, it was obtained 8 sets of gain values, each set of gain values was tested for a step input
from 1900 to 2280 ppmv and was selected one of the best gain set to study the control
response to different perturbation of H2S inlet concentration. In all these experiments the set
point value was 100 ppmv of H2S.
2.4. Effect of inlet perturbations on H2S removal efficiency.

Four perturbations were tested based in changes on inlet H2S concentration and set point
value:
- Peak of H2S inlet concentration: The inlet H2S concentration was raised from 1900 to 3800
ppmv during 1 minute at constant EBRT of 117 s (IL from 79.8 to 159.6 gS m –3 h–1).
- Stairs sinusoidal H2S inlet change: The next sine function was used: y=400sin(π/12·t) +
2800. This function was choosing to simulate the H2S inlet concentration along the day in a
wastewater treatment plant. The period was 24 h (48 stairs in each period), the mean H2S
value was 2800 ppmv and the amplitude 400 ppmv.
- Staircase steps: simulating variable IL conditions due to H2S inlet concentration changes
from 710 to 3564 ppmv at EBRT of 117 s. The period time was 14 h. This staircase steps have
been previously study by an aerobic biotrickling filter (López et al. 2016).
- H2S set-point changes: at constant inlet concentration of 1900 ppmv, the H2S set-point was
changed, up and down, between 25-200 ppmv. The H2S set point was maintained until H2S
output and nitrate dosing were stabilized.
In the first three perturbations the CV set point was 100 ppmv.
90

In this section, the results with different tuning methods are presented, also results achieved
for each control strategies tested. All experiments were performed in triplicate.
3.1. Tuning methods.

Figure 2 shows the response to the step inlet. The parameters (K, T and L) were obtained
plotting the outlet H2S concentration versus time (Figure 2). The K, T and L values were 20.7
ppmv, 113.0 min and 82.1 min, respectively. Thus, the proportional gain, integral time and
derivative time are calculated. Then, the derivative (KD) and integral gains (KI) were calculated
using the following equations:
KI=KP/I (1)
KD=KP·D (2)
Figure 2. S-shaped with step response tuning parameters K, L and T.
Table 3 show the gain values for both response method (Ziegler Nichols and AMIGO). On the
other hand, experimental results for maintained oscillation tuning method are shown in
Figure 3. The ultimate gain was 0.015 and TU was 343 s, the gain parameters calculated are
shown in Table 3.
Table 3. Gains parameters for step response and maintained oscillation methods.
Ziegler-Nichols AMIGO Maintained oscillation
Gains PID PI P PID PI PID PI P
KP 7.95·10–2 5.96·10–2 6.62·10–2 3.99·10–2 1.66·10–2 9.00·10–3 6.75·10–3 7.50·10–3
KI 5.00·10–4 2.00·10–5 - 4.00·10–4 2.00·10–5 5.24·10–5 2.36·10–5
KD 3.35 - - 1.37 - 3.86·10–1
91
Figure 3. Responses for set point with different values of KP.
As we can see in Table 3, eight set of gain values were obtained. P, PI and PID parameters
using step response tuning methods (Ziegler-Nichols and AMIGO) were closed values (same
order of magnitude). However, for maintained oscillation method the KP value was ten times
lower and the KD value was very low. So, it is expected a soft adjustment of the CV under
maintained oscillation method. However, only the AMIGO and maintained oscillation tuning
methods were able to maintain the CV at stable value. The results for the five set of gain
values are show in Figure 4 and an example of uncontrolled tuning (Figure 4(a)-PID by Ziegler-
Nichols). The system tuned by the Ziegler-Nichols step response method for P, PI and PID
was certainly unstable and a huge swing amplitude in the CV and MV.
Table 4. Comparative between tuning methods.

AMIGO Maintained oscillation
Parameter PID PI PID PI P
Offset (ppmv) 0 0.2 1.1 1.7 37.1
Settling time (h) 0.25 0.29 0.41 0.72 0.36
[H2S] mean value (ppmv) 100 99.8 98.9 98.3 137.1
[H2S] oscillation (ppmv) 93.7-107.1 94.5-105.2 93.8-104.4 94.5-102.7 131.9-142.3
Table 4 shows the offset, settling time (time to reach the set-point value), H2S concentration
mean value and H2S concentration oscillation (minimum and maximum values once reach the
set-point) for the five set of gains. As it is widely known, the main advantage of proportional
control is that only is need to adjust one parameter. However, the main disadvantage is that
the CV is operated with an offset (Smith et al. 1985). Figure 4(f) gives an example this behavior
using maintained oscillation method where the offset was of 37.1 ppmv. For the rest of
experiments, the offset was negligible (lower than 1.7 ppmv). Moreover, the mean outlet H2S
concentration value was very closed to the set-point value from 98.3 (PI by maintained
oscillation method) to 100 ppmv (PID by AMIGO method). So, this four gain values are able
to adjust the CV to the set-point value. Nevertheless, there are little differences between them,
related to settling time (Table 4) and the frequency of H2S concentration and nitrate flow rate
values (Figure 4).
92
The settling time for AMIGO method was 0.25 and 0.29 h for PID and PI, respectively and for
the maintained oscillation method was of 0.41 and 0.72 h, for PID and PI, respectively. So, it
could be expected that AMIGO method is a better method for obtain the gain values. However,
we decided to choose the PID by maintained oscillation method to carry out the effect of inlet
perturbations on H2S removal efficiency (section 3.2) due to the smoothing adjustment. Under
PID by maintained oscillation method (Figure 4(d)) there were not found almost oscillation in
the outlet H2S concentration and nitrate flow rate and the settling time was lower than PI under
the same tuning method. Otherwise, AMIGO method shows significant oscillation for PI and
PID (Figure 4 (b) and (c)).
Figure 4. Outlet H2S concentration (CV), the step of inlet H2S concentration and nitrate flow rate (MV)
for each tuning method. PID Ziegler-Nichols step response (a); PID AMIGO method (b); PI AMIGO
method (c); PID maintained oscillation (d); PI maintained oscillation (e); P maintained oscillation (f).
3.2. Effect of inlet perturbations on H2S removal efficiency (RE).

Figure 5 shows three inlet perturbations on the BTF: peaks of H2S inlet concentration,
stairs sinusoidal of H2S inlet concentration and staircase steps of H2S IL. In Figure 5(a) we
can see two H2S concentration picks from 1900 to 3800 ppmv, in both cases the same results
were obtained. In Figure 5(b) is shown a zoom of the first H2S pick, the maximum H2S
93
concentration was of 172 ppmv and, the set-point was reached in 248 s. The delay time
between the maximum H2S inlet concentration and the maximum H2S outlet concentration
was of 74 s. Similar time to reach the set-point value of 201 s, has been observed by Singh et
al. (2013), using a continuous stirred tank bioreactor (CSTR) and Ziegler-Nichols tuning
method. So, the control system is able to damp a punctual inlet H2S concentration maintained
the RE higher than 95.5%.
Figure 5. Perturbations tested with PID maintained oscillation parameters. H2S peaks (a) and (b);
Sinusoidal H2S inlet change (c); Staircase steps (d). Dotted red: outlet H 2S concentration; Dotted
blue: nitrate flow rate; black line: inlet H 2S concentration (a,b,c) and H2S IL (d).
In the stairs sinusoidal of H2S inlet concentration study (Figure 5(c)) was possible to
keep the H2S outlet concentration between 86.2 and 129.0 ppmv (mean H2S concentration of
100 ppmv). It is worth noting that oscillation in the nitrate flow rate was between 0.40-0.84 L
h–1 and therefore the N:S ratio (mol N-NO3– mol S-H2S–1 removed) was from 0.9 to 1.5. The
H2S RE was between 87.5-98.3% (mean value of 95.5%).
In the H2S staircase steps study (Figure 5(d)) the H2S outlet concentration was between
67.8-165.6 ppmv (mean outlet H2S concentration of 101.6 ppmv) and the nitrate flow rate was
between 0.13-0.76 L h–1 (N:S ratio from 0.6 to 1.2). The H2S RE was between 94.9-97.1%
(mean value of 96.4 %).
So, the results shown in Figure 5(c) and 5(b) are quite similar, despite of the differences
in the inlet profile and period time (24 and 14 h for stair sinusoidal of H2S concentration and
staircase steps of H2S IL, respectively). The N:S was within the usually operational range in
BTF without feedback control. For instance, Almenglo et al. (2016) carried out experiments at
N:S ratio between 0.34-1.74 reaching sulfate production of 8 and 95%, respectively.
The last perturbation was the modification of the H2S set-point (Figure 6). The feedback
control allows fixing any H2S outlet concentration between 25-200 ppmv under a constant H2S
inlet concentration of 1900 ppmv (EBRT of 117 s, H2S IL of 79.8 gS m–3 h–1).
Previously studies carried out, with BTFs, has been used manual and programed nitrate
dosage (Fernández et al. 2014). The manual procedure is not feasible for industrial
applications and the programmed nitrate feeding was feasible only for biogas without a change
in the H2S IL because the nitrate feeding system cannot detect the change in the H2S IL. The
feedforward control allowed the nitrate dosage at variable H2S IL. The use of a feedforward
control was proposed by Fernández et al. (2014). Another method for nitrate dosage was
carried out by Almenglo et al. (2016), the nitrate solution was fed in discontinuous mode using
the ORP measurement as a control variable. The main disadvantage of this method was the
94
outlet H2S increasing when the ORP reach the set-point due to that in this moment the nitrate
concentration in the recirculation medium was depleted.
Figure 6. Set-point changes with PID maintained oscillation parameters.
In conclusion, feedforward control using a PID tuned by maintained oscillation method allow
to damp perturbations in the H2S inlet concentration. The PID can be easily implemented in
the electric panel using a PLC with one analog input (H2S concentration) and one analog
output (variable-frequency drive of the nitrate pump). Moreover, it is needs a gas sample
conditioner cabinet with an electrochemical H2S sensor. The cost can be around 3000 € for
the PLC and 3.800 € for the H2S measurement. It has not been considered the cost of the
variable-frequency pump, nitrate pump, pH control, etc. because they are needed for a
conventional anoxic BTF without feedforward control.
4. CONCLUSIONS
The nitrate dosage in anoxic BTF is a very important matter due to its affect to the operational
cost and the H2S RE. In this study has been carried out the tuning of feedforward control using
P, PI and PID controls by three tuning methods: Ziegler-Nichols, Approximate M-constrained
Integral Gain Optimization (AMIGO) and maintained oscillation. Ziegler-Nichols step response
method, for P, PI and PID, showed not control action. On the contrary, PI and PID controllers
tuning by maintained oscillation and AMIGO methods were able to control the CV and MV at
stable values. More specifically, AMIGO method showed significant oscillation for PI and PID
in CV and MV, not so for maintained oscillation method (PI and PID controller).
Notwithstanding, PID by maintained oscillation was select to study the effect of several
perturbations over the H2S RE because there were not found almost oscillation and the settling
time was lower than using PI control.
In the study of perturbations, it is remarkable that it was possible to reduce the effect of several
perturbations in the H2S inlet concentration. For instance, H2S RE was 87.5-98.3% and 94.9-
97.1% for stairs sinusoidal and staircase steps H2S inlet concentration. It was also able to
reach a fast response to a peak perturbation and controller was able to operate satisfactorily
to a change in CV set-point. Consequently, feedback control is a good alternative to traditional
methods carried out previously for nitrate dosage such us: manual dosage, programmed
dosage or controlled by OPR.
95
ACKNOWLEDGMENTS
The Spanish Government has provided financial support to carry out the research work
through the project CTM2012-37927-C03/FEDER “Monitoring, modelling and control towards
the optimization of anoxic and aerobic desulfurizing biotrickling.
REFERENCES
Aguilar R, Poznyak A, Martın ́ ez-Guerra R, Maya-Yescas R (2002) Temperature control in

catalytic cracking reactors via a robust PID controller. Journal of Process Control 12(6):
695–705.
Almenglo F, Ramírez M, Gómez JM, Cantero D, Gamisans X, Dorado AD (2015) Modeling
and control strategies for anoxic biotrickling filtration in biogas purification. Journal of
Chemical Technology & Biotechnology. Doi: 10.1002/jctb.4769
Almenglo F, Ramírez M, Gómez JM, Cantero, D (2016) Operational conditions for start-up
and nitrate-feeding in an anoxic biotrickling filtration process at pilot scale. Chemical
Engineering Journal 285: 83–91.
Araki M (2002) PID Control. Control Systems. Robotics and Automation 2: 1–23.
Åström KJ, Hägglund T (2004) Revisiting the Ziegler–Nichols step response method for PID
control. Journal of Process Control, 14: 635–650.
Cervin A, Eker J, Bernhardsson B, Årzén KE (2002). Feedback–feedforward scheduling of
control tasks. Real-Time Systems, 23: 25–53.
Fernández M, Ramírez M, Gómez JM, Cantero D (2014) Biogas biodesulfurization in an
anoxic biotrickling filter packed with open-pore polyurethane foam. Journal of
Hazardous Materials 264: 529–535.
Fernández M, Ramírez M, Pérez RM, Gómez JM, Cantero D (2013) Hydrogen sulphide
removal from biogas by an anoxic biotrickling filter packed with Pall rings. Chemical
Engineering Journal 225: 456–463.
Hägglund T, Åström KJ (2002) Revisiting The Ziegler-Nichols Tuning Rules for PI Control.
Asian Journal of Control 4(4): 364–380.
Janssen AJH, Sleyster R, van der Kaa C, Jochemsen A, Bontsema J, Lettinga G (1995).
Biological sulphide oxidation in a fed-batch reactor. Biotechnology and Bioengineering,
47(3): 327–333.
Kleerebezem R, Mendez R (2002) Autotrophic Denitrification for Combined Hydrogen Sulfide
Removal from Biogas and Post-Denitrification. Water Science and Technology, 45
(10): 349–356.
Kramer KE, Rose-Pehrsson SL, Hammond MH, Tillett D, Streckert HH (2007) Detection and
classification of gaseous sulfur compounds by solid electrolyte cyclic voltammetry of
cermet sensor array. Analytica Chimica Acta, 584(1): 78–88.
Lee SC, Hwang YB, Chang HN, Chang YK (1991) Adaptive control of dissolved oxygen
concentration in a bioreactor. Biotechnology and Bioengineering, 37(7): 597–607.
López LR, Dorado AD, Mora M, Gamisans X, Lafuente J, Gabriel D (2016) Modeling an
aerobic biotrickling filter for biogas desulfurization through a multi-step oxidation
mechanism. Chemical Engineering Journal, 294: 447–457.
Montebello AM, Baeza M, Lafuente J, Gabriel D (2010). Monitoring and performance of a
desulphurizing biotrickling filter with an integrated continuous gas/liquid flow analyser.
Chemical Engineering Journal, 165(2): 500–507.
Montebello AM, Fernández M, Almenglo F, Ramírez M, Cantero D, Baeza M, Gabriel D (2012)
Simultaneous methylmercaptan and hydrogen sulfide removal in the desulfurization of
biogas in aerobic and anoxic biotrickling filters. Chemical Engineering Journal, 200-
202: 237–246.
Rajinikanth V, Latha K (2010) Identification and Control of Unstable Biochemical Reactor.
International Journal of Chemical Engineering and Applications, 1(1): 106–111.
96
Ramírez M, Fernández M, Granada C, Le Borgne S, Gómez JM, Cantero D (2011) Biofiltration

of reduced sulphur compounds and community analysis of sulphur-oxidizing bacteria.
Bioresource Technology, 102(5): 4047–4053.
Ramos I, Fdz-Polanco M (2014) Microaerobic Control of Biogas Sulphide Content during
Sewage Sludge Digestion by Using Biogas Production and Hydrogen Sulphide
Concentration. Chemical Engineering Journal, 250: 303–311.
Redondo R, Machado V, Baeza M, Lafuente J, Gabriel D (2008) On-line monitoring of gas-
phase bioreactors for biogas treatment: hydrogen sulfide and sulfide analysis by
automated flow systems. Analytical and Bioanalytical Chemistry, 391(3): 789–798.
Rodriguez G, Dorado A. D, Fortuny M, Gabriel D, Gamisans X (2014) Biotrickling filters for
biogas sweetening: Oxygen transfer improvement for a reliable operation. Process
Safety and Environmental Protection, 92(3): 261–268.
Syed M, Soreanu, G, Falletta P, Béland M (2006) Removal of hydrogen sulfide from gas
streams using biological processes- A review. Canadian Biosystems Engineering, 48:
2.1 –2.14.
von Meien OF, Luz Jr LFL, Mitchell DA, Ricardo PCJ, Agosin E, Fernández-Fernández M,
Arcas JA (2004) Control strategies for intermittently mixed, forcefully aerated solid-
state fermentation bioreactors based on the analysis of a distributed parameter model.
Chemical Engineering Science, 59(21): 4493–4504.
Ziegler JG, Nichols NB (1942) Optimum Settings for Automatic Controllers. Transactions of
ASME 64: 759–768.
Askesson M, Hagander P (1999) A gain-scheduling approach for control of dissolved oxygen
in stirred bioreactors. Proceedings of the 14th World Congress of International
Federation of Automatic Control. July 5-9. Beijing, China, pp 505–510.
Chotteau V, Hjalmarsson H (2009) Tuning of Dissolved Oxygen and pH PID Control
Parameters in Large Scale Bioreactor by Lag Control. Proceedings of the 21st Annual
Meeting of the European Society for Animal Cell Technology (ESACT). June 7-10.
Dublin, Ireland, pp 327–330.
Eriksson LM, Johansson M (2007) PID Controller Tuning Rules for Varying Time-Delay
Systems. Proceedings of the 2007 American Control Conference (ACC ’07). July 11-
13. New York City, USA, pp 619–625.
Hang CC, Åström KJ, Ho WK (1991). Refinements of the Ziegler-Nichols tuning formula. IEE
Proceedings D - Control Theory and Applications, 138: pp 111-118.
Singh A, Sharma V (2013) Concentration control of CSTR through fractional order PID
controller by using soft techniques. Fourth International Conference on
Communications and Networking Technologies (ICCCNT). July 4-6. Tiruchengode,
India. IEEE-31661
Ramírez M, Gómez JM, Cantero D (2015) Biogas: Sources, Purification and Uses. In Sharma
UC, Kumar S, Prasad R, Govil JN (Eds.) Hydrogen and Other Technologies Vol. 11.
Studium Press LLC, pp 296–323.
Astrom KJ, Hgglund T (1995) PID Controllers: Theory, Design, and Tuning. Instrument Society
of America, United States.
Datta A, Ho MT, Battacharyya SP (2000). Structure and synthesis of PID controllers. Springer-
Verlag London Ltd, UK.
O’Dwyer A. (2009). Handbook of PI and PID controller tuning rules (Vol. 57). Imperial College
Press, UK.
Smith CA, Corripio AB (1985) Principles and practice of automatic process control (Vol. 2).
Wiley New York.
97
Application of MBR to convert H2 and CO2/biogas

into bioCH4
Natalia Alfaro1, Israel Díaz2, Fernando Fdz-Polanco3 and María Fdz-Polanco4*
1Department of Chemical Engineering and Environmental Technology, Escuela de Ingenierías
Industriales, Sede Dr. Mergelina, University of Valladolid, Dr. Mergelina s/n, 47011 Valladolid, Spain.
e-mail: nataliaalfaro@iq.uva.es
e-mail: israel.diaz@iq.uva.es
e-mail: ffp@iq.uva.es
*Corresponding author, e-mail: maria@iq.uva.es
ABSTRACT
The bioconversion of CO2 (or biogas) into biomethane in a methanogenic bioreactor according
to reaction CO2 + 4H2  CH4 + 2H2O can create a synergy between renewable energies. On
the one hand, H2 production from wind and solar power through water electrolysis is proposed
to circumvent the limitations of intermittency and site-specificity associated to these sources.
On the other hand, by upgrading biogas with an external source of H2 from wind or solar power,
the content of bioCH4 is increased, in contrast to commercial technologies that only separate
CO2. Thus, renewable energy in the form of bioCH4 can be stored, distributed through the gas
grid or employed as fuel for vehicles. The application of membrane bioreactor with a working
volume of 60 L to convert H2 and CO2 to bioCH4 by methanogenic archaea was studied in
thermophilic conditions. A ceramic tubular membrane module was placed in the bioreactor to
generate gas bubbles, thus favouring gas-liquid mass transfer and overcoming the low
solubility of H2 in water. The loading rate was between 10 and 30 m3H2/m3reactor·d. The highest
methane yield founded was 0.22 m3CH4/m3H2, close to the maximum stoichiometric value
(0.25 m3CH4/m3H2) thus indicating that archaeas employed almost all H2 transferred to
produce CH4. Energy consumption for biogas upgrading with this technology is lower in
comparison to commercial technologies such as pressure-swing adsorption or water
scrubbing when H2 is produced from surplus renewable energies.
Keywords: biomethane; biogas upgrading; MBR; hydrogenotrophic archaea
1. INTRODUCTION
A major concern on environmental protection is the emission of greenhouse gases connected
with climate change, mainly the release of CO2 and other heat-trapping gases from human
activities (NRC, 2010). Therefore it is necessary a decrease in the CO 2 emissions which can
be achieved by reducing the amount of CO2 produced, by managing the utilization of CO2 or
the storage and fossilization of this gas (Yang et al., 2008). Actually, only the development of
mitigation technologies can decrease these emissions. In this way, a mitigation technology
that gained attention recently within the countries of the European Union is the autotrophic
conversion of CO2 and H2 to biomethane by means of methanogenic archaea because it
allows them to reduce their reliance on natural gas imports (EurObserver, 2014). It is a
biological alternative where CO2 is fixed by means of the chemoautotrophic conversion of it
98
gas and H2 to biomethane according to the reaction: 4H2 + CO2  CH4 + 2H2O (eq. 1).
Although this technology is still undeveloped because most of the H2 production worldwide
comes from stream reforming of CH4 (Ullman, 2000), it is increasing the interest on it specially
in the actual context of renewable energies implementation.
The bioconversion of CO2 (or biogas) into biomethane in a methanogenic bioreactor can
create a synergy between renewable energies. On the one hand, H2 production from wind and
solar power through water electrolysis is proposed to circumvent the limitations of
intermittency and site-specificity associated to these sources (Levene et al., 2007). There are
some moments during the day when renewable energies such as wind and solar power have
an excess of electricity production that in most of cases is lost. H2 obtaining from water
electrolysis in times of these excess electricity production from wind and solar power is a good
way of taking advantage of these renewable energies when they are in surplus because
otherwise this energy will be squander. It is an important and remarkable point nowadays in
the idea of environmental conservation and responsible use of energy. Furthermore a high
storage volume is required because of the low density of H2 and the technology for
transportation and direct utilization of this diatomic gas so the idea of direct transformation into
biomethane and the possible injection into natural gas grids or employed as fuel for vehicles
is remarkable attractive and profitable. On the other hand, anaerobic digestion of wastes is a
well-stablish renewable energy technology where the biogas obtained has a typical content of
60% CH4 and 30%CO2. This biogas can be upgraded by means of hydrogenotrophic archaeas
and an external source of H2 from wind or solar power and then, the content of biomethane
will be increased, in contrast to commercial technologies of biogas upgrading that only
separate CO2. Thereby, renewable energy in the form of biomethane can be stored, distributed
through the gas grid or employed as fuel for vehicles.
Two different approaches are showed in literature for this purpose. Firstly, the direct addition
of H2 to conventional anaerobic digesters of organic matter with the aim of the CO2 removal
from biogas while increasing the production of biomethane (Luo et al., 2012; Wang et al.,
2013; Luo and Angelidaki, 2013). Secondly, the supply of H2 and CO2 (or biogas) to an
exclusively methanogenic bioreactor rich in methanogenic archaeas (Burkhardt and Busch,
2013; Kim et al., 2013; Peillex et al., 1990). The main limitation for the technology in both
approaches is the low solubility of H2 in water (dimensionless Henry’s constant = cg/cl = 55 at
55ºC). This is also the barrier found to the successful development of the technology in an
industrial scale. To overcome this limitation, the utilization of membrane bioreactor (MBR) was
evaluated in this study to convert H2 and CO2 to bioCH4 in thermophilic conditions. A ceramic
membrane module was employed to create a large gas sparging surface and the feasibility of
the technology was assessed.
2.1. Pilot plant

One insulated cylindrical MBR with a working volume of 60 L was fed H2 and CO2 from gas
cylinders and the rate was regulated with mass-flow controllers. Feed line was mixed with the
recirculation and then preheated in a thermostatic bath at 55 ºC. A ceramic tubular membrane
module (ATECH, Germany, 5 L volume, approximately area of 1 m2 and 0.8 µm pore size)
was used as gas sparging surface to generate small bubbles. Its upper part was connected
with the feed line as shown in Figure 1. The reactor counted with a compressor to recirculate
99
biogas from the reactor´s headspace through the membrane module. A peristaltic pump was
employed to avoid solids deposition at a rate of 1000 mL/min.
Figure 20. Diagram of the MBR studied in the experiment.
2.2. Operating conditions

Anaerobic sludge from a thermophilic anaerobic digester treating activated sludge from
WWTP of Valladolid was used to inoculate the reactor. A set-up period was performed by
supplying H2 and CO2 in a ratio of 4:1 (according to the stoichiometric values of eq. 1) at a
loading rate of 5.0 m3H2/m3reactor·d with a gas recirculation rate of 11.6 m3/d for 17 days in
thermophilic conditions (all values expressed at 55ºC and 1 atm). Afterwards, the experiment
started and was divided in three stages according to Table 1. Each one corresponds to a
certain gas load rate with the objective of determining the maximum loading rate that could be
applied with a 95% conversion efficiency for methane. Different recirculation rates were
applied in order to evaluate mass transfer conditions and reactor performance.
Table 6. Operating conditions studied during the experiment.

Stage 2 Stage 2
Stage 1 Stage 3
2a 2b
t (d) 0 25 81 148
Loading rate, LR
10 20 20 30
(m3H2/m3reactor·d)
Recirculation
rate, QR 8.0 8.0 12.3 12.3
(L/min)
Nutrients required for microbial activity and a phosphate buffer solution were supplied when
the concentration of NH4+ and PO43- fell below 500 mg/L, macronutrients and micronutrients
were added too. The macronutrient solution was prepared like the stock solution A reported
in Angelidaki and Sanders (2004), while the micronutrients solution was a version that was
modified (by adding 500 mg/L of resazurine) from the trace-metal solution also from Angelidaki
and Sanders (2004) and the phosphate buffer solution was prepared with K2HPO4·3H2O and
KH2PO4 to a final pH of 7.4 like in Diaz et al. (2015).
100
2.3. Monitoring and analysis

Temperature was maintained at 55 ±1ºC. Headspace pressure was monitored with a Cerabar
PMC131 probe of Endress Hauser. A gas flowmeter was employed to measure the effluent
gas rate (by liquid displacement) and the composition of its biogas was determined by gas
chromatography (GC-TCD) as described in Diaz et al. (2010). Liquid effluent was collected
and measured in a graduate cylinder. pH, TSS, VSS and NH4+ values were analyzed
according to Standard Methods (APHA, 2005), VFA concentration was analyzed by gas
chromatography (GC-FID) following the method reported in Alcántara et al. (2014).
Calculations about efficiency of H2 utilization, methane yield, mass flow rate of H2 transferred
from gas to liquid phase, effluent mass flow rate of CH4 gas as equivalent H2, kLaH2 and kLaCO2
values, maximum specific utilization rate and fraction of H2 employed for methanogen growth
have been performed following calculations in Diaz et al. (2015).
3.1. Conversion of H2 and CO2 to CH4

Biomass adaptation to the substrate took place during the set-up period when the feed mass
flow rate of H2 gas (mH2,IN) used was 25.2 g/d (LR of 5.0 m3H2/m3reactor·d) and the QR 8.0 L/min.
A large part of the H2 fed in these first days was transferred to the liquid phase and consumed
but was not employed for CH4 production, probably due to biomass adaptation to the
substrate. Then, first stage started with a mH2,IN of 49.9 g/d and the same recirculation rate than
set-up period. The mass balance performed to the gas phase showed that the average
efficiency of H2 utilization (ƞH2) was 95% and an average methane yield (YCH4) of 0.18
m3CH4/m3H2 was observed. Average mass flow rate of H2 transferred from gas to liquid phase
(mH2, GL) obtained was 44.6 g/d and an effluent mass flow rate of CH4 gas as equivalent H2
((mCH4, OUT)H2eq) according to eq.1 in average of 35.0 g/d.
On day 25, mH2,IN was raised to 99.9 g/d while QR was maintained at 8.0 L/min (stage 2a). The
increase in the mass flow rate provoked a slightly decrease in average ƞH2 until 85.7% thus
indicating that mass transfer conditions were still acceptable even when the LR was doubled.
However, the average YCH4 obtained was slightly higher than in the previous stage, 0.19
m3CH4/m3H2. The difference between mH2, GL and (m CH4, OUT)H2eq was almost exactly the same in
stage 1 and 2a and the values obtained were 82.2 g/d and 72.5 g/d respectively. Biogas
recirculation rate was increased to 12.3 L/min with the purpose of raising ƞH2 and the stage 2b
started. Under this conditions, the performance of the MBR improve significantly, reaching an
average ƞH2 value of 95% and YCH4 reached of 0.21 m3CH4/m3H2, the highest obtained up to
then. In this case, mH2, GL observed was 89.2 g/d and 80.1 g/d of (m CH4, OUT)H2eq. The difference
between them was in the same range than the others. This values were somewhat higher than
the ones obtained in stage 2a, thus indicating that recirculation improved the amount of H2
transferred from gas to liquid phase and the amount of CH4 produced.
At this point, mH2,IN was augmented to 149.8 g/d in combination with a maintained QR of 12.3
L/min (stage 3). During this stage, ƞH2 was 95% in average while YCH4 increased until 0.22
m3CH4/m3H2, much closer to the maximum stoichiometric value of 0.25 m 3CH4/m3H2. The
difference between mH2, GL and (m CH4, OUT)H2eq was lower than in previous stages meaning that
archaeas employed almost all H2 transferred in order to produce CH4.
101
The MBR successfully transformed at least 95% of the H2 fed at LR between 9.6 and 28.8
m3H2/m3reactor·d adjusting the gas recirculation rate. This highest LR is similar than that achieved
on a similar pilot-scale bioreactor (Diaz et al., 2015) with a hollow fibre membrane module and
higher than those found in packed column bioreactors (4.5 m3H2/m3reactor·d) (Burkhardt and
Busch, 2013) or CSTR (18 m3H2/m3reactor·d) (Kim et al., 2013).
3.2. Mass transfer capacity in the MBR

The average kLaH2 values observed during the different stages in the experiment are shown in
the Table 2. The kLaH2 values obtained ranged from 88 h-1 (for the lowest total gas flow through
the membrane) to 268 h-1 (for the highest). The estimated kLaCO2 values were from 63 h-1 to
190 h-1.
Table 2. Average kLa obtained in the different stages during the experiment of H2 and CO2.
Stage 2 Stage 2
Stage 1 Stage 3
2a 2b
kLaH2 (h-1) 88 92 157 268
kLaCO2 (h-1) 62 65 111 190
It should be draw attention to this maximum kLaH2 value is higher than others found in
bioreactors with traditional gas diffusers (at equivalent gas rates) and in the range of CSTR
with high agitation speeds as 700 rpm (Kreutzer et al., 2005). This is the result of the large
sparging area of the membrane module employed, which produces a good gas-liquid mass
transfer area.
3.3. Biological Activity

It is very important the fact that the adaptation of an unspecific anaerobic thermophilic sludge
to H2 and CO2 has been accomplished. As a result, a methanogenic archaeas population
developed, which was capable of the bioconversion of H2 and CO2 into bio CH4. From an
industrial point of view, it can be translated into lower acquisition costs of specific
hydrogenotrophic methanogens because an unspecific anaerobic sludge could be used as
inoculum. The maximum specific utilization rate (U) and the fraction of H2 employed for
methanogen growth (fx) are shown in Table 3.
Table 3. Average fraction of H2 employed for methanogen growth in the different stages during the
experiment of H2 and CO2.
Stage 2 Stage 2
Stage 1 Stage 3
2a 2b
U (gCOD/gVSS·d) 6.3 7.7 3.96 7.5
fx 0.22 0.12 0.10 0.09
The maximum specific utilization rate obtained was 7.7 gCOD/gVSS·d withinthe range of typical
design value suggested from methanogens growing on H2 and CO2 (Rittman, 2001). According
to the results shown in Table 3, it could be stated that the highest value of fx was obtained
during the first stage. This fraction has dropped along the experiment, and in the last stage a
decrease of more than 50% appeared. This fact took place because at the beginning of the
experiment, specially in the set-up period, an important fraction of H2 was utilized for microbial
102
growth but when the sludge was completely adapted to the gas substrates, only a small
fraction of H2 was used for methanogen growth, almost all H2 transferred was used to produce
CH4.
VFA concentration was very low during the experiment: acetic acid concentration was under
100 mg/L and propionic acid was below 50 mg/L. It has been observed that when a period
started and the concentration of TSS and VSS were lower than 1g/L, VFA is higher than in
other times. As a consequence, it seems to be that the pathway of H2 and CO2 to acetic acid
and later to CH4 is more important than the pathway of H2 and CO2 directly to CH4. pH was
over the experiment between 6.8-7.4. Average total and volatile suspended solids
concentration analysed during the experiment in the several stages of the experiment are
shown in Table 4. This values shown an increasing trend from Stage 1 to Stage 2b. The
operation was stopped then for few hours and the bioreactor opened in order to observe the
state of the membrane and as a result some liquid was lost and TSS and VSS decrease.
Table 4. Average Total Suspendid Solids (TSS) and Volatile Suspended Solids (VSS) obtained in the
different stages during the experiment, g/L.
Stage 2 Stage 2
Stage 1 Stage 3
2a 2b
TSS (g/L) 0.61 1.05 2.15 1.02
VSS (g/L) 0.35 0.88 2.00 0.89
3.4. Energy Consumption

Energy consumption in the process is due to heating for achieving thermophilic conditions,
power for biogas recirculation and energy so as to obtain H2. Among this three process
consuming energy, the production of H2 is the largest, with a fix amount of 7.2 kwh/m3 biogas.
Energy consumption add up to 8.4 kw·h/m3 biogas for the lowest LR and 7.7 kw·h/m3 biogas
for the highest LR studied in the experiment as shown in Figure 2.
Taking the advantage of H2 obtaining from water electrolysis in times of excess electricity
production from wind and solar power when they are in surplus, the bioconversion processes
of H2 and CO2 such as the studied will be energetically beneficiaries, its energy consumption
can be decreased to a great extent. Taking into account that H2 can be obtained from water
electrolysis using surplus electricity production from renewable energies, energy consumption
during the biological upgrading performed in this study was around 0.6-0.47 kw·h/m3 biogas
at LR between 20-30 m3H2/m3reactor·d (Figure 2). This value is slightly higher than others
obtained (Diaz et al., 2015) with hollow fibre membrane modules (0.2-0.3 kw·h/m3 biogas)
because of the highest pressure drop within the ceramic module. However, fortunately, this
energy consumption is lower than the average value of 0.7 kw·h/m3 biogas found on the most
used commercial technologies such as pressure-swing adsorption or water scrubbing (SCG,
2013).
103
Energy Consumption
9
Consumption (kwh/m 3 biogas) 8
7
6
Heating + Biogas
5 Recirculation+ H2
4 production
3
Heating + Biogas
2
Recirculation
1
0
0 10 20 30 40
H2 IN rate (m 3 H2/m 3reactor·d)
Figure 2. Energy Consumption in the process.
4. CONCLUSIONS
The bioconversion of H2 and CO2 into bioCH4 was feasible using an unspecific anaerobic
thermophilic sludge as an inoculum after an adaptation period. The maximum loading rate of
30 m3H2/m3reactor·d had a 95% efficiency in H2 utilization and a methane yield of 0.22
m3CH4/m3H2, close to the stoichiometric maximum achievable of 0.25 m3CH4/m3H2. Gas
sparging through the ceramic module membrane showed a high capacity of H2 mass transfer
in the operating conditions studied. Taking the advantage of H2 can be obtained in water
electrolysis process using surplus electricity production from wind and solar power, energy
consumption of the process was lower than the value found on the most used commercial
technologies.
ACKNOWLEDGMENTS
This research was supported by the Spanish Ministry of Education, Culture and Sports
(FPU13/04680 Grant) because human resource and the funding company FCC-Aqualia in the
project Smart Green Gas.
REFERENCES
Alcántara C, Fernández C, García-Encina P, Muñoz R (2014) Mixotrophic metabolism of

Chlorella sorokiniana and algal-bacterial consortia under extended dark-light periods and
nutrient starvation. Applied Microbiology and Biotechnology 99:2393–2404.
APHA, AWWA, WPCF (2005) Standard Methods for the Examination of Water and
Wastewater. 21st ed. Washington, DC.
Burkhardt M, Busch G (2013) Methanation of hydrogen and carbon dioxide. Applied Energy
111:74–79.
104
EurObserver (2014) Biogas barometer. http://www.eurobserv-er.org/biogas-barometer-2014/

Accessed October 2015
Diaz I, Pérez C, Alfaro N, Fdz-Polanco F (2015) A feasibility study on the bioconversion of

CO2 and H2 to biomethane by gas sparging through polymeric membranes. Bioresource
Technology 185: 246-253.
Kim S, Choi K, Chung J (2013) Reduction in carbon dioxide and production of methane by
biological reaction in the electronics industry. International Journal of Hydrogen Energy
38: 3488–3496.
Kreutzer MT, Kapteijn F, Moulijn JA, Ebrahimi S, Kleerebezem R, Van Loosdrecht MCM
(2005) Monoliths as biocatalytic reactors: smart gas–liquid contacting for process
intensification. Industrial and Engineering Chemistry Research 9646–9652.
Levene JI, Mann, MK, Margolis RM, Milbrandt A (2007) An analysis of hydrogen production
from renewable electricity sources. Solar Energy 81:773–780.
Luo G, Angelidaki I (2012) Integrated biogas upgrading and hydrogen utilization in an

anaerobic reactor containing enriched hydrogenotrophic methanogenic culture.
Biotechnology and Bioengineering 109:2729–2736.
Luo G, Angelidaki I (2013) Co-digestion of manure and whey for in situ biogas upgrading by
the addition of H2: process performance and microbial insights. Applied Microbiology and
Biotechnology 97:1373–1381.
NRC, The National Research Council (2010) Advancing the Science of Climate Change. A
report from The National Academies. Washington, DC.
Peillex J, Fardeau M, Boussand R, Navarro J, Belaich JP (1988) Growth of Methanococcus

thermolithotrophicus in batch and continuous culture on H2 and CO2: influence of
agitation. Applied Microbiology and Biotechnology 29:560–564.
Rittman B (2001) Environmental Biotechnology: Principles and Applications. McGraw Hill,

New York.
Ullman F (2000) Ullmann’s Encyclopedia of Industrial Chemistry. 7th ed. Wiley, Weinheim.
Wang W, Xie L, Luo G, Zhou Q, Angelidaki I (2013) Performance and microbial community
analysis of the anaerobic reactor with coke oven gas biomethanation and in situ biogas
upgrading. Bioresource Technology 146:234– 239.
Yang H, Xu Z, Fan M, Gupta R, Slimane RB, Bland AE, Wright I (2008) Progress in carbon
dioxide separation and capture: a review. Journal of Environmental Science 20:14–27.
105
Application of feedback and feedforward control

strategies in aerobic biotrickling filters for biogas
desulfurization
Luis R. López1, Mabel Mora1, Caroline Van der Heyden2, Eveline Volcke2, Juan A.
Baeza1, Javier Lafuente1, David Gabriel1*
1 GENOCOV Research Group, Department of Chemical, Biological and Environmental Engineering,

Escola d’Enginyeria, Universitat Autònoma de Barcelona, 08193 Bellaterra, Spain. *Corresponding
author: Email: David.gabriel@uab.cat
2 Department of Biosystems Engineering, Ghent University, Coupure links 653, 9000 Ghent, Belgium
Email: Eveline.Volcke@UGent.be
ABSTRACT
The H2S loading rate (LR) in biotrickling filters (BTFs) varies due to diurnal and seasonal
biogas flow and H2S inlet ([H2S]in) concentration variations, thus affecting the BTF
performance in terms of H2S removal efficiency (RE) and sulfate selectivity. The airflow rate
(AFR) and the trickling liquid velocity (TLV) constitute potential control handles (manipulated
variables) to increase the stability of BTFs performance. Still, process control is not often
implemented in industrial BTFs. In this contribution, BTFs stability against [H2S]in
concentration and biogas flowrate changes was assessed both experimentally and through a
model-based analysis. Feedback controllers based on airflow regulation and O2
measurements versus TLV regulation coupled to H2S measurements were put forward. The
controller stability and controllability limits were assessed through model-based analysis.
Then, feedback and feedforward control loops were implemented separately in a lab-scale
BTF operating under real LR fluctuant conditions. Experimental evaluation of feedback control
strategies to control [H2S]in concentration variations through TLV regulation showed to be an
adequate strategy to face fluctuating LR due to [H2S]in concentration changes. However,
controlling biogas flowrate changes at an EBRT below 69 s presented a highest difficulty using
classical controllers, since the process in under mass transfer limitation. A feedforward control
loop based on TLV regulation showed as a better strategy to reduce load perturbations impact
and more importantly, to improve sulfate selectivity.
Keywords: desulfurizing biotrickling filter; biogas; modeling; feedback and feedforward control
strategies; O2 transfer improvement.
1. INTRODUCTION
Biogas desulfurization using BTfs represents a significant technological innovation that serves
to improve the quality of biogas produced by anaerobic digestion (AD) in wastewater treatment
plants (WWTPs) to be further used for energy recovery in downstream processes such as
electricity production. Process optimization can be achieved through application of control
strategies. Control strategies in BTFs are needed in order to have a robust and reliable
desulfurization process that allows obtaining a gaseous effluent with certain characteristics
independently of the operational conditions. Process control studies in BTFs is almost null if
compared to other biological processes, from technologies with a lower implementation level
such as composting treatment (Puyuelo et al., 2010) up to the most advanced and developed
technologies like wastewater treatment in WWTPs (Guerrero et al., 2011) and AD (Nguyen et
al., 2015). However, lately more and more interest has been placed in understanding the
effect of the main manipulated variables involved in the control of biogas desulfurization on
106
BTFs (López et al., 2016a; Rodriguez et al., 2014, 2013). Moreover the degree of knowledge
of this technology has allowed the calibration and validation of models for BTFs for biogas
desulfurization under aerobic conditions (Bonilla-Blancas et al., 2015; López et al., 2016b) as
well as under anoxic conditions (Almenglo et al., 2015). These last efforts denote a serious
intent of researchers to optimize technology performance through the development of control
strategies, which is still required to warrant the stability of long-term operation. The main
restriction for obtaining a stable long-term operation of BTFs for biogas desulfurization under
aerobic conditions, is the elemental sulfur (S0) accumulation inside the BTF packed bed due
to insufficient oxygen (O2) mass transfer efficiency at high LR (López et al., 2016a, Rodriguez
et al., 2014), therefore control strategies must be focused in this sense.
Control strategies are needed in BTFs for biogas desulfurization in order to mitigate process
disturbances. In full-scale BTFs, both [H2S]in concentration and biogas flowrate can vary
simultaneously. On one hand, H2S]in concentration increase may occur when the organic
matter content of the feedstock to an anaerobic digester increases (Syed et al., 2006). On the
other hand, biogas flowrate may vary in facilities with small gasholders for biogas storage or
when high amounts of biogas are produced due to an increase in the feedstock to AD (Nguyen
et al., 2015; Walsh et al., 1988). In a desulfurizing BTF, the effect of LR variations provoked
by [H2S]in increase or decrease is that the stoichiometric O2/S ratio is moved from complete
sulfide oxidation (H2S→SO42-; O2/S≥2) to partial sulfide oxidation (H2S→S0;O2/S≤2), the latter
leading to S0 production. Therefore, a control system is needed in order to prevent
environmental and technical problems when unexpected LR variations occur due to [H2S]in
concentration changes or due to biogas flowrate fluctuations. AFR increase is one of the most
extended actuations in industrial facilities to provide additional O2 to avoid massive S0
accumulation (Rodriguez et al., 2014). Nevertheless, this strategy results in biogas dilution
and to a consequent loss of its heat capacity. For this reason, strategies based on the
improvement of O2 transfer from the gas to the liquid phase are key to optimize the complete
oxidation of H2S to sulfate (López et al., 2016a). The objective of the strategies would be to
search the enhancement of the desulfurization process without diminishing biogas calorific
power, e.g. due to an increase in the air supply flowrate. Strategies based on the regulation
of the TLV can provide interesting results in terms of sulfate selectivity improvement since
increasing the TLV improves both the dissolved oxygen load (DOL) and the penetrability of
the liquid through the BTF bed, thus reducing possible dissolved oxygen (DO) gradients inside
the bed without the abovementioned biogas dilution (López et al., 2016a). Regulation of TLV
also modifies the global mass transfer coefficient (KLa.). The wettability factor and the liquid
hold-up are the governing factors for the effect of TLV.
The aim of this work was to evaluate, both experimentally and through modelling of a BTF,
the control limits and capabilities of different control strategies such as feedback and
feedforward control loops applied to biogas desulfurization in aerobic BTFs to control the
composition of the upgraded biogas. The added value of the different control strategies in
terms of sulfate selectivity were also quantified. The simulation study also aimed to help to
plan better experiments and designing control strategies in the lab-scale reactor. Simulation
of control strategies for BTFs for H2S removal that properly improves reactor behaviour during
different operational scenarios (LR increments due to concentration increase or biogas
flowrate increments) has a great importance since these studies are lacking in literature.

2.1. Model-based analysis of Feedback control strategies
107
A model-based analysis was performed to evaluate the stability and controllability limits of the
feedback control strategies under study. To this end, a BTF model describing accurately
physical-chemical and biological processes for the removal of high loads of H2S from biogas
streams was applied (López et al., 2016b). The BTF model describes the reactor performance
in terms of H2S removal efficiency and predicts satisfactorily the main intermediate (S0) and
final product produced during biological oxidation of H2S (Mora et al., 2016). The performance
of the different control strategies in terms for H2S removal efficiency (RE), O2 consumption, S0
and sulfate (SO42-) production was assessed through simulation. In this work, the analytical
approach through the use of time-integral performance criteria was used for controllers tuning.
Time-integral criteria were selected for controllers tuning, since they are based on the entire
response of the process, unlike the simple criteria that use only isolated characteristics of the
dynamic response. The aim during control tuning was to determine the controller parameters
that lead to reduce time-integral criteria while producing also a stable, not oscillatory response.
The most used are Integral of the Square Error (ISE), Integral of the Absolute Error (IAE) and
Integral of the Time-weighted Absolute Error (ITAE) as described in Eq. 1. to Eq. 3,
respectively.
∞
ISE= ε2 (t)∙ dt (1)
0
∞
IAE= ε(t) dt (2)
0
∞
ITAE= t∙ ε(t) dt (3)
0
Where ε is the error, determined as the difference between the setpoint (SP) and the
measured variable. However, results regarding the selection of the most suitable parameters
for each control strategy are not shown because the aim of this study was to show the
application of different control strategies and their added value in aerobic BTFs for biogas
desulfurization. Two case were studied in order to evaluate feedback control strategies in a
BTF for biogas desulfurization when LR was varied from the reference LR of 56.3 g S-H2S m-
3 -1
h up to a LR of 168.9 g S-H2S m-3 h-1 due to changes in [H2S]in concentration or biogas
flowrate. LR conditions were also tested under open-loop conditions in order to quantify
process improvement when control loop was implemented. The first case evaluated was the
effect of changes in the [H2S]in concentration, which was step-wise increased from 2000 ppmv
up to 6000 ppmv as shown in Fig.1a.
a) b)
250
8000 225
7000 200
Qbiogas (L h-1)
6000 175
Cg,H2S (ppmv)
150
5000
125
4000
100
3000
75
2000 50
1000 25
0 0
0 200 400 600 800 1000 1200 0 200 400 600 800 1000 1200
Time (h) Time (h)
Figure 1. Inlet profile to simulate (a) [H2S]in concentration changes and (b) biogas flow rate changes.
108
Each concentration was evaluated long enough to achieve steady-state conditions as well as
to study the long-term effect of the disturbances. The second case evaluated was the effect
of changes in the biogas flowrate performing step-wise increments from 61.5 L h-1 (EBRT=118
s) to 184.5 L h-1 (EBRT=48.4 s) as it can be observed in 1b. In this contribution, two feedback
and one feedforward control strategies were studied to cope with loading rate variations
caused by changing influent H2S concentrations or biogas flow rate. The first strategy
concerns airflow regulation Closed-Loop (AFR-CL, Fig 2a), which is the traditional control
strategy applied in BTFs (Rodriguez et al., 2014). This strategy controls the O2 concentration
in the outgoing air, [O2]out, by manipulating the air flowrate: an increasing LR results in an
increase of O2 consumption and therefore a need of a larger air supply. A SP of [O2]out
concentration of 5.70 % (v v-1) was established in AFR-CL.
a)
11
Cg,H2Sin 1
Qbiogas
4
6
3
Cg,O2,out,M Cg,O2,out,SP
7 12
CT
13
ε
8
2 10 FC 9
b)
11
Cg,H2Sin 1
Qbiogas
4
6
3
Cg,H2S,out,M Cg,H2S,out,SP
7 12
CT
13
ε
8
2 10 FC 9
Figure 2 . Schematic of control strategies in a co-current BTF set up. (a) Airflow regulation Closed-
Loop (AFR-CL). Control of [O2]out through air flowrate. (b) TLV regulation Closed-Loop (TLV-CL).
Control of [H2S]out through liquid recirculation flowrate. (1) biogas inlet , (2) Air flow inlet to the aeration
column, (3) aeration column, (4) Air flow inlet to the BTF, (5) main reactor, (6) biogas outlet, (7) liquid
recycled, (8) H2S and O2 transmitters, (9) feedback flow controller, (10) recycling pump, (11) recycling
liquid line to the BTF (12) mineral medium and bicarbonate supply, (13) liquid purge. Q biogas
109
Regarding to the air flowrate range, this was regulated from the reference case value of 24.2
L h-1 to a maximum value of 120.8 L h-1 corresponding to a dilution factor of biogas of 28.2 %
and 66.3 %, respectively. Considering that methane (CH4) composition in biogas can be in the
range of 50 to 70% (v v-1) (Walsh et al., 1988), a minimum CH4 content in air of 16.9 % would
be obtained when the above mentioned dilution factor is applied. According to Walsh et al.
(1988), this value is higher than 15% which is the upper explosive limit (UEL), defined as the
maximum concentration of CH4 required to obtain an explosive mixture. Such gas flow rates
and corresponding biogas dilutions are nonsense from a practical application perspective.
However, the analysis from a theoretical perspective was considered as a worth exercise to
be done.
The second strategy studied, namely TLV regulation Closed-Loop (TLV-CL), is an innovative
strategy to control biogas desulfurization in BTFs. A detailed description of the TLV-CL control
strategy is presented in Fig. 2b). This control loop aimed at controlling the [H2S]out
concentration through liquid recirculation flowrate regulation when LR is varied due to [H2S]in
concentration or biogas flowrate changes. In TLV-CL, the final control element is the liquid
recirculation pump, which regulates the liquid recirculation flowrate that is related to TLV. A
SP of [H2S]out concentration of 100 ppmv of H2S was established to calculate the error. This
SP was selected to produce a high-quality biogas effluent to enhance life time of co-generation
engines or the engine where biogas is finally burned. Regarding TLV regulation limits, a
minimum TLV of 5.9 m h-1 was applied when the [H2S]out concentration was below the SP in
order to provide nutrients and moisture to the packed bed. A maximum TLV of 28.3 m h-1 was
set to prevent BTF bed flooding. Table 1 summarizes the different feedback control strategies
when the different controlled variables and perturbations are combined.
Table 1 Feedback control strategies under study.

Closed- Controlle
Manipulated Disturbance Controller Paramete
Loop d Range
Variable s type r
Scenario Variable
0.005-
P Kc
0.2
Cg,H2S,in
AFR-CL [O2]out Qair 0.005-
Kc
PI 0.2
τI 1-5
Cg,H2S,in P Kc 1-1000
TLV-CL [H2S]out Qbiogas & Kc 1-1000
PI
Qbiogas τI 0.1-10
Only digital P and digital PI controllers were considered. Derivative action was not included
because during previous tests it was observed that due to the characteristics of the
disturbances tested, the anticipative effect was null since the response of the process was a
constant nonzero error, giving no action for the derivative term (dε⁄dt =0).
Sulfur species mass balance referred to the total amount of H2S at the inlet and the amount
of O2 consumed during each simulation were calculated to evaluate the added value of the
different feedback control strategies. Mass balances using each control strategy were
compared with mass balances of simulations under Open-Loop conditions. Therefore in order
to compare the BTF performance at each LR tested, mass balances were calculated by
integration of data of each LR step. Sulfur mass balances and the amount of O 2 consumed
were determined according to Eq.4 to Eq. 7:
110
g S-SO2-
4,produced (4)
% SO2-
4,produced = ∙ 100 (%)
g S-H2 Sin
g S-S0produced (5)
% S0produced = ∙ 100 (%)
g S-H2 Sin
g S-H2 Soutlet (6)

% H2 Soutlet = ∙ 100 (%)
g S-H2 Sin
% g O2,consumed =(g O2,inlet -g O2,outlet ) ∙ 100 (%) (7)
Therefore once the percentage of each specie and the O2 consumed was determined, it was
compared with the values obtained during the Open-Loop case as described on Eq. 8 in order
to determine the % of change (improvement or reduction).
%Open-Loop- %Closed-Loop (8)

% Change= (%)
%Open-loop
2.2. Application of feedback control strategies based on TLV regulation
Two case studies were here analysed in order to evaluate feedback control in a lab-scale BTF
for biogas desulfurization when LR is varied due to [H2S]in concentration or due to biogas
flowrate changes. Further details about the experimental setup and analytical methods can be
found on López et al. (López et al., 2016a). Additionally to the BTF set up, a digital controller
was programmed. The digital controller determined the controller action according to Equation
9 (Baeza, 2016).
𝑛
∆𝑡 𝜏𝐷
𝐶𝑛 = 𝐶𝑠 + 𝐾𝐶 [𝜀𝑛 + ∑ 𝜀𝑘 + (𝜀𝑛 − 𝜀𝑛−1 )] (9)
𝜏 ∆𝑡
𝑘=1
The first case studied was the evaluation of a feedback control loop in order to face changes
in the [H2S]in concentration. Table 2 shows the main characteristics of the first profile tested in
the BTF. Tests were named according to the control strategy used, the disturbance applied,
namely concentration (C) or biogas flowrate (EBRT), and according to the average LR of the
test. Accordingly, the abovementioned test was defined as feedback-C-154 (FB-C-154). The
procedure followed to perform FB-C-154 was as follows. First, the BTF performance was
assessed with fluctuant [H2S]in concentration conditions at constant TLV (5.9 m h-1). Then, the
same LR profile was applied together with a feedback control strategy. In agreement with
section 2.1, a [H2S]out SP concentration of 100 ppmv was defined. TLV was limited in the range
5.9 to 28.3 m·h-1 when used as manipulated variable.
Table 2. Experimental conditions for feedback control evaluation under variable [H2S]in concentration
Constant Variable
[H2S] LRavg [H2S]avg [H2S]max [H2S]min
Test TLV test TLV test
(g S-H2S m-3 h-1) (ppmv) (ppmv) (ppmv) (m h-1) (m h-1)
FB-C-154 154 5460 5770 5283 5.9 5.9 - 13
111
The second case evaluated was the effect of biogas flowrate changes performing stepwise
increments from 150 L h-1 (EBRT=67.4 s) to 166 L h-1 (EBRT=60.5 s). According to the
terminology above defined, this test was named feedback-EBRT-120 (FB-EBRT-120). The
main characteristics of FB-EBRT-120 profile tested in the BTF are described in table 3. A
similar procedure to that performed in the FB-C-154 test was performed for the FB-EBRT-120
test. First, the BTF performance was assessed with fluctuant biogas flowrate profiles at a
constant TLV (5.9 m h-1). Then, the same biogas flowrate profile was applied together with a
feedback control strategy.
Table 3. Experimental conditions for feedback control evaluation under variable biogas flowrate
Constant Variable
[H2S] LRavg [EBRT]avg [EBRT]max [EBRT]min
Test TLV test TLV test
(g S-H2S m-3 h-1) (s) (s) (s) (m h-1) (m h-1)
FB-EBRT-120 120 64.9 67.4 60.5 5.9 5.9 - 17
The duration of the FB-C-154 and FB-EBRT-120 profiles was 1.3 hours (80 minutes) and 4
cycles were performed for each test. Since the effect of TLV over S0 and sulfate selectivity in
the long-run has already been assessed and proved to be efficient (López et al., 2016a),
analysis in this work targeted the RE solely as process performance variable.
2.3. Application of a feedforward control strategy based on TLV regulation
Variable LR profiles to evaluate the performance of a feedforward control loop were designed
and tested in order to simulate daily LR fluctuations and elimination capacities (EC) commonly
found in industrial facilities. The first case studied was the evaluation of a feedforward control
loop in order to face LR variations due to [H2S]in concentration changes. Feedforward control
strategy consisted of manipulating the TLV as a function of the [H2S]in (Fig. 3a).
a) b) 2D Graph 9
160 40
Cg,H2Sin
Qbiogas 140 LR 35
TLV
120 30
LR (g S-H2S m h )
-1
-3
TLV (m h-1)
100 25
Feedfoward 80 20
controller
60 15
40 10
Cg,H2S,out,M
20 5
0 0
0 2 4 6 8 10 12 14
Time (h)
Figure 3. (a) Schematic of feedforward control strategy for controlling LR variations due to [H 2S]in
concentration through regulation of liquid recirculation flowrate and (b) LR inlet profile (solid line) and
TLV profile (dashed line) designed to control variable LR conditions due to [H2S]in concentration for the
FF-C-78 case.
Experimental results obtained in López et al. (2016a) were useful to determine the relation
between the DOL and TLV, since to have an optimal feedforward controller is important to
know the effect of the manipulated variable over the process. In consequence in this section
a DOL/[H2S]in SP of 0.32 g O2 g-1 S-H2Sin was set, since experimentally it was found 100%
112
sulfate selectivity (S-SO42-/S-H2Sremoved) and H2S removal could be reached at this ratio. To
simulate variable LR conditions due to [H2S]in concentration changes, three different cycles
of 14, 18 and 20.5 hours were designed for each test and continuously repeated during the
test period (≈200 h). The main characteristics of the different LR profile tested with a
feedforward control loop are presented in table 4.
Table 4. Experimental conditions for feedforward control loop evaluation under variable [H2S]in
concentration
Constant Variable
[H2S] LRavg [H2S]avg [H2S]max [H2S]min
Test TLV tests TLV tests
(g S-H2S m-3 h-1) (ppmv) (ppmv) (ppmv) (m h-1) (m h-1)
FF-C-78 78 2759 5000 1000 6.5 - 19
FF-C-98 98 3459 7000 1000 6.5 6.5 - 30
FF-C-116 116 4119 8500 1000 6.5 - 37
Tests were named according to the disturbance applied and the average LR of the test. e.g.
for the test with an average LR of 78 g S-H2S m-3 h-1 the test was named FF-C-78. The
procedure followed to perform these tests was as follows. First, the BTF performance was
assessed under open-loop with fluctuant LR conditions at constant TLV (6.5 m h-1) following
the profile described in Fig.3b (black solid line). Then, the same LR profile was applied
together with a feedforward control strategy, regulating TLV in function of the [H2S]in as shown
in Fig. 3b (black dashed line). It must be mentioned that, before each test, the BTF was
operating during 8 days at a pseudo steady-state in order to study the dynamics of the process
in each test from similar starting conditions. Afterwards, the sulfate selectivity was determined
as the total amount of sulfate produced per amount of H2S removed from the gas phase (S-
SO42-/S-H2Sremoved) along the whole test with and without feedforward control. Then, in order
to evaluate the regulation limits of TLV along the three tests, sulfate selectivity increase for
each test was determined as the increase of sulfate selectivity when the process was
controlled related to the sulfate selectivity under Open-Loop conditions.

3.1. Feedback control strategies evaluation to face variable LR inlet changes
The results of the actuation of the controller and the [H2S]out concentration for a P and PI
controller in TLV-CL strategy to face [H2S]in concentration changes are presented in figure 4.
a) b)
22
700
20 Kc 147 I 1
600 SP Kc 300
Open Loop 18
[H2S ]out (ppmv)
500 Kc 147 I 1 16
TLV (m h-1)
Kc 300
400 14
12
300
10
200
8
100 6
0 4
0 200 400 600 800 1000 1200 0 200 400 600 800 1000 1200
Time (h) Time (h)
Figure 4. (a) [H2S]out (controlled variable) and (b) trickling liquid velocity (manipulated variable) for H2S
control facing inlet H2S concentration changes using a P and PI controller with TLV-CL strategy. SP
=set point.
113
Fig. 4a shows that a significant reduction of the [H2S]out offset down to values close to only 20
ppmv when a Kc of 300 was used for the P controller. Different controller parameters for the P
and PI controller were also tested along the controller tuning procedure, however here only
two cases using the best set of parameters are shown. It has to be noticed that lower controller
response was obtained when the PI controller was applied (Fig 4b), which was translated to a
higher offset compared to the P controller. At the end of the second concentration step, an
offset of only 20 ppmv was obtained with the P controller, while with the PI controller an offset
of 40 ppmv was obtained. The reason why the same response was always obtained and,
therefore, the [H2S]out reached was always the same independently of the parameters tested
in the PI controller can be explained by the characteristics of the error during the last [H2S]in
concentration step. Once the error was constant, the change of the proportional action tend to
zero and the PI controller turned into an integral (I) controller.
Regarding the controllability of the process under the conditions tested, Fig. 4a indicates that
the process is controllable at the highest LR tested of 168 g S-H2S m-3 h-1, also indicates that
no controller action is needed before reaching that LR. This procedure was also followed to
evaluate controllability and stability of AFR-CL control strategy, applying the same types of
disturbances and evaluating different controller parameters (as described in section 2.1) in
order to determine the most suitable conditions for this strategy (data not shown). Besides
considering the control limits of the different control strategies, it was important to analyse the
added value of the different feedback strategies on biogas desulfurization in an aerobic BTF.
In order to have additional criteria to select control strategies, it was important to evaluate if
the strategy studied, besides achieving the control objective (e.g. reaching the SP value),
improved process performance significantly. A proper H2S removal is necessary in full-scale
BTFs for biogas desulfurization. However, a large product recovery as sulfate instead of S0 is
a key factor to ensure process stability in the long run. In Table 5, the improvement in the main
process variables for AFR-CL and TLV-CL control strategies facing [H2S]in concentration
changes using P and PI controllers are presented.
Table 5. Improvement on the main process performance variables compared to the process under
Open-Loop conditions, for AFR-CL and TLV-CL facing [H2S]in concentration changes using a
feedback P and PI controller.
Improvement Reduction Reduction Improvement
Cg,in H2S Control Manipulated g O2
S- SO42-produced S-S0produced g H2S outlet
type variable consumed
(ppmv) (%) (%) (%) (%)
4000 P 0.4 4.5 1.6 0.3
AFR
6000 Kc=0.1 1.7 3.8 14.1 2.1
4000 PI 0.5 6.2 2.5 0.4
Kc=0.074 AFR
6000 τI =1 3.1 8.5 23.7 3.0
P
6000 Kc=300 TLV 17.5 91.0 75.6 9.0
PI
6000 Kc=147 TLV 14.9 70.7 71.9 8.7
τI =1
Table 5 shows that the O2 control through AFR regulation reduces the amount of H2S in the
outlet and the amount of S0 produced because of a larger O2 consumption that leads to an
increase in the amount of sulfate produced if compared to the process under Open-Loop
conditions. However, reduction and improvement percentages are much lower compared to
114
those obtained with the TLV-CL control strategy. The AFR-CL control strategy showed less
effectivity than TLV-CL control strategy since controlling [O2]out concentration does not ensure
a better gas-liquid mass transfer, which is the main limiting step of the desulfurization process.
Through TLV regulation, O2 consumption was also substantially increased without any biogas
dilution effect if compared to the AFR-CL control strategy. According to López et al. (López et
al., 2016a), TLV regulation improves the penetrability of the liquid through the packed bed
reducing possible DO gradients and, therefore, the formation of S0.
Regarding the comparison between P and PI controllers, a PI control law performed better in
the AFR-CL control strategy because the PI controller was able to remove completely the
offset, thus improving the O2 transport and consumption onto the packed bed. However, the
P controller provided larger improvements in the main performance variables than the PI
controller in the TLV-CL control strategy. As stated previously this fact was mainly due to the
type of disturbance employed in this study that turned the PI controller into an I controller. In
any case, the TLV-CL control strategy offered interesting results in order to improve process
performance since a reliable BTF operation can be obtained through the regulation of TLV,
which does not involve biogas dilution.
As it was abovementioned, AFR-CL was included in this study to have a more complete
theoretical analysis, however results observed in Table 5 for AFR-CL indicate that this strategy
barely improves main process performance variables, therefore further results corresponding
to this strategy are not showed. In figure 5 the result of the actuation of the controller and the
[H2S]out for a P and PI controller in TLV-CL strategy to face biogas flowrate changes is
presented. Contrarily to the previous cases when TLV was regulated in order to control LR
increase due to [H2S]in concentration increases, the Open-Loop response shows that the
[H2S]out concentration was already higher than the SP during the first step change. Thus, the
first biogas flowrate step provoked an actuation in the controller output (Fig 5b) denoting that
the process was much more sensitive to the effect of reducing the EBRT (biogas flowrate
increase) than to changes in the [H2S]in concentration. Interestingly, at the highest LR tested
(169 g S-H2S m-3 h-1), the [H2S]out concentration was reduced from 1100 ppmv (Open-Loop
response) down to 550 ppmv, which was far from the SP (100 ppmv) meaning that changes in
the LR due to a biogas flowrate increase corresponding to the highest LR tested were not
controllable through a TLV-CL control strategy.
a) b)
30
1200
1100 25 Kc 110 I 1
1000 SP Kc 300
Open Loop
900
[H2S ]out (ppmv)
Kc 110 I 1 20
TLV (m h-1)
800
Kc 300
700
15
600
500
400 10
300
200 5
100
0 0
0 200 400 600 800 1000 1200 0 200 400 600 800 1000 1200
Time (h) Time (h)
Figure 5. (a) [H2S]out concentration (controlled variable) and (b) TLV (manipulated variable) for H 2S
control facing biogas flowrate changes using a P and PI controller.
115
However, the [H2S]out was on the SP during the second LR tested (112.6 g S-H2S m-3 h-1),
which means that under those LR and EBRT the process was controllable. The
uncontrollability of the process during the last LR step was due to a limitation on the gas-liquid
mass transfer when the EBRT was reduced. Therefore, mass transfer of gaseous compounds
such as H2S and O2 is affected. Gas-liquid mass transfer could be improved by using intensive
G-L mass transport devices such as Jet Venturi in order to produce an intensive contact
between gas and liquid (Rodriguez et al., 2014). In this sense, model-based analysis of control
strategies denotes their usefulness, since results such as the one shown in Fig. 5 warn to the
designer that under those LR and EBRT conditions a redesign of the BTF would be needed
in order to achieve higher H2S transfer conditions. In the other hand, Table 6 shows the
improvement in the process variables considered using P and PI control laws in a TLV-CL
control strategy to face biogas flowrate changes.
Table 6. Improvement on the main process performance variables for TLV control strategy facing
biogas flowrate changes using a using a P and PI feedback controllers
Control Manipulated Improvement Reduction Reduction Improvement
Qbiogas
type variable S- SO42-produced S-S0produced g H2S outlet g O2 consumed
(L h-1) (%) (%) (%) (%)
P
123 TLV 22.4 85.6 53.0 7.3
Kc=300
PI
123 Kc=110 TLV 15.23 51.52 43.22 6.67
τI =1
As above discussed, the process was out the controllable limits using TLV as manipulated
variable when the highest biogas flowrate was tested. Therefore, such results were excluded
from the analysis herein. Results in Table 6 show that the TLV regulation improved the main
process performance variables during variable biogas flowrate conditions. Despite the lower
EBRT of operation compared to the Open-Loop case, H2S RE and O2 consumption were
improved due to an enhancement of mass transfer along the packed bed caused by the TLV
regulation. This results are in agreement with López et al. (2016c), where TLV increase
improved the global mass transfer coefficients (KLa) for H2S and O2, which correlates well with
the reduction of the [H2S]out and on the increase of O2 consumption. Consequently, product
selectivity towards sulfate was improved.
The main difference between the P and PI control laws were related to the TLV applied in
each case. If a P controller was used, a maximum TLV of 21.9 m h-1 was used by the controller
while a maximum TLV of 15.1 m h-1 was reached with the PI controller during the first biogas
flowrate step. Again, differences in the performance between both controllers was conditioned
to the characteristics of the disturbance used in this work. Irrespective of the type of controller,
process performance was positively improved through TLV regulation in order to face biogas
flowrate changes. However, controlling biogas flowrate changes at an EBRT below 69 s can
hardly be done using classical controllers in this setup.
116
3.2. Experimental evaluation of feedback control facing changes in the [H2S]in

concentration
For the FB-C-154 test, comparison of the [H2S]out concentration profile between the Open-
Loop and Closed-Loop case is presented in Fig. 6a, while the LR and the controller response
(TLV) is presented in Fig. 6b.
a) b)
150 180 24
175 22
H2S-LR (g S-H2S m-3 h-1)

125 20
170
Cg,out,H2S (ppmv)
18
TLV (m h-1)
100 165 16
75 160 14
155 12
50 10
150
8
25
145 6
0 140 4
0.0 0.5 1.0 1.5 2.0 2.5 3.0 3.5 4.0 4.5 5.0 5.5 0.0 0.5 1.0 1.5 2.0 2.5 3.0 3.5 4.0 4.5 5.0 5.5
Time (h) Time (h)
Figure 6. Experimental concentration profiles of Feedback control facing [H 2S]in concentration

variations for test FB-C-154: (a) [H2S]out concentration profiles under Open-Loop conditions (black line)
and under Closed-Loop conditions (dark gray line) and SP value (back dashed line) and (b) LR profile
(black line), TLV actuation under Open-Loop conditions (black dashed line) and controller actuation
(TLV) (dark gray line).
Fig. 6a shows that the [H2S]out concentration of the process under Open-Loop conditions
exceeds the setpoint during LR peaks. Under Closed-Loop conditions, [H2S]out concentration
remains below the setpoint instead. A significant reduction of the [H2S]out concentration error,
down to 19.8 ppmv at time 0.75h, was obtained for the process under Closed-Loop conditions,
while an error up to 48.9 ppmv was obtained at time 4.75 h, during the Open-Loop case.
Regarding the controller actuation, a stable controller response was obtained; TLV varied
between 5.9 m h-1 and 13.3 m h-1. Also it is important to remark the fact that in biological
processes control, such as in BTFs for biogas desulfurization, the adaptability of the process
to operational conditions must be taken into account and, therefore, the need of use of different
set of controller parameters. When LR profiles were tested under Closed-Loop conditions
along several hours, even during some days (data not shown), the processes increases its
capability and therefore the parameters selected are less effective.
In the case of FB-C-154 test, a difference in [H2S]out concentration along the time from the first
to the third peak can be observed due to this phenomena. When the BTF is operated under
Closed-Loop conditions, regulation of TLV increased the amount of water retained in the BTF
packed bed, consequently increasing the H2S absorption. This is why the error is higher under
Closed-Loop conditions during the valley periods of the LR profile. On the other hand, a
difference in the [H2S]out concentration between the first three cycles and the fourth cycle is
observed, which could be explained by a change in the preferential paths of the gaseous and
liquid phase, causing a reduction of the removal efficiency (RE). In spite of these phenomena,
which are common in multiphase reactors such as BTFs, the controller was able to control
[H2S]out concentration below the setpoint during the last cycle.
117
One particularity of the digital controller of our process is that there is a controller actuation
even when the process does not exceeds the setpoint under Closed-Loop conditions. This
was explained because of the digital controller programming (Eq. 4). The digital controller
considers the previous control action and the difference of errors between the present error
(εn ) and the previous error (εn-1 ). Therefore always that this difference of errors is different
of zero, and the previous controller actuation is higher to the minimum control action, there is
a change in the controller’s output. As it is shown in Fig. 6b and 7b, TLV for the Closed-Loop
never returned to the minimum TLV (red dashed line).
3.3. Experimental evaluation of feedback control facing changes in the biogas flowrate
A comparison of the [H2S]out concentration profiles between the Open-Loop and Closed-Loop
case for the FB-EBRT-120 test is presented in Fig. 7a, while the controller response (TLV)
and the LR profile are presented in Fig. 7b.
a) b)
140 24
150
135 22
H2S-LR (g S-H2S m-3 h-1)
125 20
130
Cg,out,H2S (ppmv)
18
TLV (m h-1)
100 125 16
75 120 14
115 12
50 10
110
8
25 105 6
0 100 4
0.0 0.5 1.0 1.5 2.0 2.5 3.0 3.5 4.0 4.5 5.0 5.5 0.0 0.5 1.0 1.5 2.0 2.5 3.0 3.5 4.0 4.5 5.0 5.5
Time (h) Time (h)
Figure 7. Experimental concentration profiles of Feedback control facing biogas flowrate variations for
test FB-EBRT-120: (a) [H2S]out concentration profiles under Open-Loop conditions (black line) and under
Closed-Loop conditions (dark gray line) and SP value (back dashed line) and (b) LR profile (black line),
TLV actuation under Open-Loop conditions (black dashed line) and controller actuation (TLV) (dark
gray line).
In Fig. 7a is observed how the [H2S]out concentration for the process under Closed-Loop is
well controlled during variable biogas flowrate conditions. Instead, the [H2S]out concentration
is above the setpoint for the Open-Loop case, especially during biogas flowrate peaks. Under
Open-Loop conditions the highest [H2S]out concentration error was around 28 ppmv, while
under Closed-Loop conditions, the highest error obtained was 8.1 ppmv. Since FB-EBRT-120
test was performed at a lower EBRT (65 s) than the reference EBRT (120 s), the process was
under mass transfer limitation. Subject to this situation, variations of biogas flowrate were
much more difficult to control compared to the previous test (FB-C-154) were [H2S]in
concentration was varied. This difference can be observed in the fact that higher TLV values
in the range of 5.9-17 m h--1 were applied, indicating that higher actuations were needed to
attenuate the effect of this disturbance over the process. In addition, this mass transfer
limitation effect in the FB-EBRT-120 test, can be observed in the minimum of the LR profile
that for the Open-Loop and Closed-Loop case, the [H2S]out concentration was almost the same.
This evidenced that the TLV applied in the valley periods of the LR profile for the Closed-Loop
case was not capable to attenuate or improve the effect of the disturbance if high gas-liquid
118
flowrate relations are applied, as it occurred in the FB-C-154 test. However, TLV actuations
were effective to control [H2S]out concentration during H2S-LR peaks in the FB-EBRT-120
profile, especially during the last cycle in which the relation between the gas-liquid flowrate
was lower.
3.4. Experimental evaluation of feedforward control facing changes in the [H2S]in

concentration
Fig. 8 (a, b and c) shows the effect of feedforward control loop implementation on [H 2S]out
concentration for the LR tested (78, 98 and 116 g S-H2S m-3 h-1). In Fig. 8 is observed how for
the TLV regulated process, [H2S]out concentration is considerably lower along the experimental
time. However, long-term operation allowed to discover some interesting phenomena, such
as mass transfer resistance increase due to S0 accumulation, diminishing the superficial area
of the packing material decreasing therefore the absorption capacity. Interestingly, the profile
obtained from FF-C-98 (Fig. 8b) shows the effect of biomass growth and biomass density
increment, since [H2S]out decreases progressively along the test without any other change in
the system both in constant and variable TLV experiments.
a) 700 b) 3000
2700
600
2400
[H2S]out (ppmv)
[H2S]out (ppmv)
500 2100
1800
400
1500
300 1200
200 900
600
100
300
0 0
0 50 100 150 200 250 0 50 100 150 200
Time (h) Time (h)
c) 2000 d) 20 % Sulfate Selectivity increase
1800 100
( g S-SO4 g S-H2Sremoved )
1600 18
[H2S]out (ppmv)
1400 80
% Sulfate Selectivity
1200 16
60
1000
2- -1
800 14
40
600
400 20
Open Loop 12
Closed Loop
Selectivity increment
200
0 0 10
70 75 80 85 90 95 100 105 110 115 120
0 50 100 150 200
Time (h) LR H2S (g S-H2S m -3 h-1)
Figure 8. Concentration profiles and sulfate selectivity obtained from variable LR operations with and
without the implementation of feedforward control. (a) [H2S]out concentration for FF-C-78 experiment.
(b) [H2S]out concentration for FF-C-98 experiment. (c) [H2S]out concentration for FF-C-116 experiment.
Experiments with constant TLV (Open-Loop) are presented with solid black line and experiments with
variable TLV (Closed-Loop) are presented with solid red line. (d) Sulfate selectivity (Open-Loop: black
squares; Closed-Loop: grey triangles) and sulfate selectivity increment (white circles) obtained with and
without the implementation of feedforward control.
119
This behaviour was not observed in tests FF-C-78 and FF-C-116. This fact demonstrated the
difficulty of studying control strategies in BTFs with long-term operations, since factors like
biomass growth and S0 accumulation modifies the BTF performance, making difficult the
differentiation between the impact of disturbances and long-term effects. At the lowest average
LR (FF-C-78) the biomass concentration was probably increasing slowly, which did not
increase the EC significantly, and the adaptation of the biomass was not necessary at such
low load. At the highest average LR (FF-C-116) the problem was probably that, despite the
biomass concentration was higher after previous tests, the maximum oxidation rate was
already reached and the process was not controllable under those conditions. Regardless the
adaptation of the biomass, with the implementation of the control strategy the [H2S]out was kept
under 500 ppmv in most of the cases after the first 25h of each test. This result is especially
important since the control strategy proposed in this work avoids shock loads of H2S to energy
recovery processes which causes corrosion and damage of engines (Walsh et al., 1988).
In Fig. 8d sulfate selectivity, with and without feedforward control, is presented on the left Y
axis. It can be observed that increasing the TLV according to the LR increase led to an
enhancement of sulfate selectivity up to 92%, which corresponded to a relative sulfate
selectivity increase up to 16.5 % (right Y axis in Fig 8d). Similar results were obtained by
Rodriguez et al. (2014) who obtained an increment of 17% on sulfate selectivity using a jet-
venturi device to improve O2 transfer efficiency. This means that the implementation of a
feedforward control strategy together with the use of intensive transfer devices would improve
significantly the operation of desulfurizing BTFs while reducing operational and maintenance
costs. Also it must be noticed that sulfate selectivity for Open-Loop tests were considerably
higher compared to those values obtained when comparable step-wise LR increments were
applied for short periods of time (Montebello et al., 2010). Sulfate selectivity during Open-Loop
tests also supports the fact that a better biomass density distribution was achieved along the
BTF bed when a variable LR was applied to the BTF.
4. CONCLUSIONS
Air flowrate Closed-Loop showed to be inefficient in terms of process improvement when

added-value variables such as sulfate selectivity were considered. On the other hand, TLV
Closed-Loop showed to be a feasible strategy to improve complementary BTF performance
parameters such as removal efficiency, sulfate selectivity and O2 consumption. The type of
disturbance used to study control strategies did not allow evaluating the proportional integral
controller properly in all the cases studied. When the two manipulated variables were
compared, better results were obtained through TLV regulation since no biogas dilution is
involved and because higher percentage of process improvement is achieved.
Experimental application of a feedback control strategy showed that TLV based
strategies improved BTF stability against [H2S]in concentration changes and biogas flowrate
changes. Moreover, LR ranges experimentally studied were in agreement with controllability
ranges predicted by the model-based analysis. Interestingly, TLV regulation allowed reducing
H2S concentration peaks despite the BTF was under mass transfer limitation during biogas
flowrate changes. When a feedback control strategy was coupled to a desulfurizing BTF
operating under variable LR conditions, [H2S]out concentration was properly controlled through
TLV regulation, especially in the case when biogas flowrate was varied since the process was
under mass transfer limitation. Satisfactory results of the application of feedforward control
were thanks to a previous rigorous analysis of the effect of TLV over the process performance
under Open-Loop conditions. TLV regulation in a feedforward control loop during long-run
operation allowed, besides observing an H2S removal improvement, minimizing S0 production
by increasing O2 gas-liquid transport along the packed bed.
120
ACKNOWLEDGMENTS
The Spanish government provided financial support through the CICYT project CTM2012-
37927-C03. The Department of Chemical, Biological and Environmental Engineering at UAB
(Universitat Autònoma de Barcelona) is the unit of Biochemical Engineering of the Xarxa de
Referència en Biotecnologia de Catalunya (XRB), Generalitat de Catalunya.
REFERENCES
Almenglo, F., Ramírez, M., Gómez, M., Cantero, D., David, A., 2015. Modeling and control
strategies for anoxic biotrickling filtration in biogas purification. J. Chem. Technol.
Biotechnol.
Baeza, J.A., 2016. Principles of bioprocess control. In: Current Developments In
Biotechnology and Bioengineering. Elsevier.
Bonilla-Blancas, W., Mora, M., Revah, S., Baeza, J.A., Lafuente, J., Gamisans, X., Gabriel,
D., González-Sánchez, A., 2015. Application of a novel respirometric methodology to
characterize mass transfer and activity of H2S-oxidizing biofilms in biotrickling filter beds.
Biochem. Eng. J. 99, 24–34.
Guerrero, J., Guisasola, A., Vilanova, R., Baeza, J.A., 2011. Improving the performance of a
WWTP control system by model-based setpoint optimisation. Environ. Model. Softw. 26,
492–497.
López, L.R., Bezerra, T., Mora, M., Lafuente, J., Gabriel, D., 2016a. Influence of trickling liquid
velocity and flow pattern in the improvement of oxygen transport in aerobic biotrickling
filters for biogas desulfurization. J. Chem. Technol. Biotechnol. 91, 1031–1039.
López, L.R., Dorado, A.D., Gabriel, D., Mora, M., Gamisans, X., Lafuente, J., 2016b. Modeling
an aerobic biotrickling filter for biogas desulfurization through a multi-step oxidation
mechanism. Chem. Eng. J. 294, 447–457.
López, L.R. 2016c. Development and application of control strategies in an aerobic biotrickling
filter for H2S removal from biogas streams: experimental and modelling study. Universitat
Autònoma de Barcelona. PhD Thesis.
Montebello, A.M., Baeza, M., Lafuente, J., Gabriel, D., 2010. Monitoring and performance of
a desulphurizing biotrickling filter with an integrated continuous gas/liquid flow analyser.
Chem. Eng. J. 165, 500–507.
Nguyen, D., Gadhamshetty, V., Nitayavardhana, S., Khanal, S.K., 2015. Automatic process
control in anaerobic digestion technology: A critical review. Bioresour. Technol. 193, 513–
522.
Puyuelo, B., Gea, T., Sánchez, A., 2010. A new control strategy for the composting process
based on the oxygen uptake rate. Chem. Eng. J. 165, 161–169.
Rodriguez, G., Dorado, a. D., Bonsfills, a., Gabriel, D., Gamisans, X., 2013. Optimization of
Oxygen Transfer through Membrane Diffusers for Biological Sweetening of Biogas.
Chem. Eng. Technol. 36, 513–518.
Rodriguez, G., Dorado, A.D., Fortuny, M., Gabriel, D., Gamisans, X., 2014. Biotrickling filters
for biogas sweetening: Oxygen transfer improvement for a reliable operation. Process
Saf. Environ. Prot. 92, 261–268.
Syed, M., Soreanu, G., Falletta, P., Béland, M., 2006. Removal of hydrogen sulfide from gas
streams using biological processes - A review. Can. Biosyst. Eng. 2.1.
Walsh, J., Ross, C., Smith, M., Harper, S., Wilkins, W., 1988. Handbook on Biogas Utilization.
Georgia.
121
Oral Presentations – Algal Bacterial Systems

Tuesday June 7th 2016
122
Kinetic Exploration of the Microalgal Growth Rate of

Species Residing in a Waste Stabilization Pond
Bjorge Decostere1, Andres M Alvarado2, Esteban AM Sanchez2 , Diederik Rousseau1,
Ingmar Nopens3 and Stijn WH Van Hulle1
1
Department of Industrial Biological Sciences, Ghent University, Karel de Goedelaan 5, B-
8500 Kortrijk Bjorge.Decostere@Ugent.be
2
Departamento de Recursos Hidricos y Ciencias Ambientales, Cuenca University, Av. 12 de
Abril s/n, Cuenca – Ecuador. a.alvarodomartinez@gmail.com
3
Department of Mathematical Modelling, Statistics and Bio-informatics, Ghent University,
Coupure Links 653, B-9000 Gent. Ingmar.Nopens@Ugent.be
ABSTRACT
Since the oxygen concentration is an important operational parameter of Waste Stabilization

Ponds (WSP), accurate insight in the oxygen production kinetics of the microalgae living in
such systems, is essential in view of system optimization. In this research the growth kinetics
of the most dominant microalgal species, namely Chlorella vulgaris was investigated when
different conditions of light intensity and temperature were imposed. This microalgal species
was isolated from the WSP situated near Cuenca, Ecuador. A combined respirometric and
titrimetric set-up was used to assess the microalgal kinetics. The experimental results
illustrated the interdependent relationship of light intensity and temperature, which had a
significant influence on the microalgal growth rate. Consequently the growth kinetics,
proposed in a previous developed model were extended with a mathematical function that
describes this relationship. Additional combined respirometric and titrimetric data was used
for model calibration and model validation. Two parameters were considered for model
calibration, namely the maximum specific growth rate and the oxygen mass transfer
coefficient. Based upon the Theil’s Inequality Criterium, the model described the dissolved
oxygen production and the related proton consumption rather good.
Keywords: Microalgae; Waste Stabilization Pond; Respirometry; Titrimetry; Modelling
1. INTRODUCTION
Wastewater treatment by Waste Stabilization Ponds (WSP) is widely used in the world to treat
different types of wastewater, ranging from domestic to industrial waste water. The most
important advantage of this system is its simplicity in construction and operation. Moreover
WSP exhibits a high reliability because its operation depends mainly on biological processes
and does not depend on equipment performance that can fail (Von Sperling, 2007). In the
treatment of domestic wastewater by WSP, the aerobic stabilization of organic compounds by
bacteria and the oxygen production and nutrient removal by algal (photosynthetic) activity are
the main occurring natural processes. The oxygen demand by bacteria for the assimilation of
organic substrate is met by the oxygen produced through algal photosynthetic activity. The
photosynthetic activity of algae depends on several environmental conditions such as light
(Richmond, 2004), temperature (Bordel et al., 2009) and availability of nutrients (Yao et al.,
2010; Broekhuizen et al., 2012). Further, Carvalho and Malcata (2003) demonstrated the
interdependent relationship of light intensity and temperature. The basic assumption for this
123
was, that for a given temperature, there is a direct relation between light intensity and
activation energy and as such a light dependency of the activation energy (Carvalho and
Malcata, 2003). Also, the light saturation level is influenced by the temperature. Indeed,
augmentation of temperature shifts the light saturation point to higher light intensities and as
such also the intensity at which photo-inhibition occurs (Sorokin and Krauss, 1962).
Alvarado (2012) stated that the oxygen production and consequently the amount of dissolved
oxygen in the pond, is a fundamental operational parameter for both maintain a healthy
aerobic biomass and to induce adverse conditions in the ponds for pathogen viability. Thus, it
is essential to (accurately and frequently) quantify the influence of operational conditions over
the oxygen production. Further in view of system optimization, good insight in the kinetics of
the microalgal biomass is a prerequisite. For this, the combined respirometric and titrimetric
methodology (Decostere et al., 2013) was used in this research to determine the growth
kinetics of microalgal species in the facultative pond of the full scale installation at Ucubamba,
Ecuador. This WSP is designed to treat the domestic effluent form the city of Cuenca
(Ecuador). The combined respirometric and titrimetric methodology involves the measurement
of the dissolved oxygen production and the proton consumption at constant pH by
photosynthetic activity during a batch – wise conducted experiment. The main isolated
microalgal species in the WSP was Chlorella vulgaris and therefore this species will be used
during the experiments. In particular, the effect of light intensity and temperature was
assessed. In addition a kinetic growth model, based upon the activated sludge models (ASM)
(Henze et al., 2000) was developed, calibration and validation was performed with the specific
experimental data.
2.1. Cultivation of the microalgae

Chlorella vulgaris was isolated from the biomass of the full – scale WSP installation located
near Cuenca. Isolation was done by controlled growth on specific media. After isolation, the
microalgal species was bred axenic in continuous stirred 3.0 L reactors with ideal inorganic
carbon, nutrients and light availability. Also the reactors were periodically sparged with air to
prevent settlement of the microalgae on the walls of the breeding reactors and stripping of
oxygen produced by the microalgae. Further the microalgae were kept in exponential growth
phase by bi-weekly refreshing of the breeding medium and harvesting of microalgae.
2.2. Experimental protocol

The combined respirometric and titrimetric unit (Fig. 1) was similar to Decostere et al. (2013).
The 1.6 L reactor vessel was heat-jacketed to allow temperature control (Alpha R8,
www.lauda.de) enabling the exploration of system behaviour at different temperatures. The
light cage enclosing the reactor entirely consisted of 36 fluorescent lamps (Voltech,T5 8 W).
Light intensity was measured using a lux light meter (FC 840020, Sper Scientific).
Dissolved oxygen (DO) and pH were measured online with an oxygen (Inpro 6870i, Mettler
Toledo) and pH electrode (Inpro R 4260 i/SG/120, Mettler Toledo) and the data logged using
a PCI-MIO-16XE-50 data acquisition card using LabView (www.ni.com). The pH was
controlled online at a user defined set-point using a banded (+/- 0.1 pH) on-off feedback control
algorithm implemented in LabView by dosing HCl or NaOH through two 3-way pinch solenoid
valves (Cole Parmer). The rate and amount of 0.25 M HCl and 0.5 M NaOH dosed into the
reactor vessel constitutes the titrimetric data.
124
For the respirometric tests microalgae from the breeding reactor were used, after
centrifugation at 4000 rpm for 5 minutes (Heraeus Megafuge 8 Centrifuge, Thermo Scientific).
The microalgae where then rinsed twice with demineralized water. Subsequently, the
concentrated algae were diluted in 1.5 L of demineralized water. Next 100 ml of nutrient
solution containing nitrogen and phosphorus was spiked into the solution. Then, a sample
(200 ml) was taken to analyse the initial microalgal biomass and nutrient concentrations.
Finally, 100 ml of bicarbonate solution was spiked before the start of each test. As such, it
was strived to keep the amount of nutrients and algal biomass for each separate test
constant, namely 15 g NH4+ – N m-3, 0.6 g NO3- – N m-3, 1.5 g PO43—P m-3 and 100 g HCO3-
m-3. The microalgal biomass concentration was 100 g DW m-3 . The pH was controlled at 7.5
± 0.1 for each separate test.
Fig 1. Schematic overview of the combined respirometric – titrimetric setup.
In order to investigate the effect of light intensity and temperature and possible interaction
between those environmental variables, a 2 level full factorial design (Box and Draper, 1987)
was at first used. As such 7 experiments (Table 1) were performed. The maximal
photosynthetic activity, expressed as g O2 g DW -1d-1 was considered as
response variable (y). This value was derived from the oxygen production rate (OPR) curve
(Decostere et al., 2013), more specifically the maximum value of this curve. Next linear
regression was used to determine the significant variables and possible interaction. The
equation to describe this photosynthetic activity can be denoted as:
𝑦 = 𝑏0 + 𝑏1 𝑋1 + 𝑏12 𝑋1 𝑋2 + 𝑏2 𝑋2 (Eq.1)
With 𝑏0 , 𝑏1 , 𝑏12 and 𝑏3 the different coefficients and 𝑋1 , 𝑋2 the variables influencing the
photosynthetic activity, respectively light intensity (lux) and temperature (K). The interaction
term in this equation is represented by 𝑏12 𝑋1 𝑋2 . The statistical analysis was performed with
SPSS.

Nitrate, phosphate and ammonium were measured according to the standard methods
(APHA, 2005). The microalgal dry weight concentration was determined by turbidity
measurements, which was previously related to the microalgal dry weight by means of the
linear equation shown below:
125
𝑌 = 3.06 𝑋 − 8.54 (Eq.2)
With 𝑌 ( g DW m-3) the microalgal dry weight concentration and 𝑋 (FTU) the feature of turbidity
measurement.
2.4. Modelling software

For the simulations described in this work, the Flexible Modelling Environment (FME) package
(Soetaert and Herman, 2009) was used. Although this open access package only uses a
textual interface, recently it has been more intensively used in view of ecological modelling
(Haario et al., 2009; Mannina et al., 2012)
Parameter estimation was performed by the minimization of an objective function by using an

optimization algorithm. The objective function was defined as SSE between model prediction
and measurements, which could be denoted as:
𝐽 (𝜃) = ∑𝑧𝑗 ∑𝑁
𝑖=1(𝑦𝑖,𝑗 − 𝑦𝑖,𝑗 (𝜃))
2
(Eq.3)
in which 𝐽(𝜃) represents the objective function based on N data points and 𝑦𝑖,𝑗 and 𝑦𝑖,𝑗 (𝜃)
represent the measured data and predicted value for the variable 𝑗 respectively, with 𝑖 the
index of the data points ranging from 1 to N. 𝜃 includes the parameters to be estimated. It can
be derived from this equation, that the objective function calculation was performed for the
variables (here dissolved oxygen concentration and proton addition) simultaneously.
Minimization of the objective function was done by the Levenberg – Marquardt (Yu and
Wilamowski, 2011) algorithm.
2.5. Goodness of fit

When model calibration and validation was performed, the goodness – of – fit between
measured and calculated was quantified by calculating the Theil’s inequality coefficient (TIC)
(Theil, 1961) which can be denoted as follows:
√∑𝑖(𝑦𝑖 − 𝑦𝑖,𝑚 ) 2
𝑇𝐼𝐶 = (Eq. 4)
√∑𝑖 𝑦𝑖2 + √∑𝑖 𝑦𝑖,𝑚
2
in which 𝑦𝑖 represents simulated data and 𝑦𝑖,𝑚 represents measured data points. A TIC value
lower than 0.3 (Coppens et al., 2014) thereby indicates a good agreement with measured
data.
3.1. Determination of significance of factors

To determine the significance of light intensity and temperature, seven separate experiments
were run. The relative maximum photosynthetic activity of each experiment was considered
as respons variable for the linear regression. This was derived from the respirometric profiles
(Table 1). Results of the linear regression (95 % confidence level) showed the significance of
the interdependency of light intensity and temperature (p – statistic = 0.024). The
interrelationship between temperature and light intensity can be explained by the fact that
there is a direct relation between light intensity and activation energy. Furthermore, the light
saturation level is influenced as mentioned before by the temperature, next to the prevailing
126
light intensity (Carvalho and Malcata, 2003). As such this aspect was taking into account for
model development.
Table 1. Initial settings of light intensity and temperature and corresponding relative maximum
photosynthetic activity for C. vulgaris
I T 𝑝𝑂2,𝑚𝑎𝑥
(lux) (K) (g O2 g-1 DW d-1)
4810 283 0.08
4810 306 0.59
4810 290 0.13
10650 290 0.12
1000 290 0.26
10650 298 0.66
10650 303 1.00
3.2. Model development

In the previous section, experiments were performed to assess the effect of light intensity and
temperature and the interdependent relationship between those variables. In this section, a
model taking into account inorganic carbon, phosphorus and nitrogen kinetics (Decostere et
al., 2016) was extended. This allows to describe several experiments conducted at different
temperatures and light intensities by using one uniform combination of parameter values
The interrelationship between temperature and light intensity was described here as (Carvalho
and Malcata, 2003):
𝐼
𝐾1 𝐼 −𝛽( )
𝑓 (𝐼, 𝑇) = (𝐾 )𝑒 𝑇 (Eq. 5)
2 𝑇+𝐼
Where 𝐼 (lux) and 𝑇 (K) represent respectively the prevailing light intensity and temperature.
ß (-) represents a constant related to the activation energy and ideal gas constant. 𝐾1 (-) and
𝐾2 (K-1) are constants related to the light intensity and temperature.
The growth rates proposed by Decostere et al. (2016) on different inorganic carbon sources
and nutrients can subsequently be extended as follows:
𝑆𝐶𝑂2 𝑆𝑁𝐻+ 𝑆𝑃𝑂3− 𝐾1 𝐼 −𝛽 𝐼

𝜌𝐴𝑙𝑔(𝐶𝑂2,𝑁𝐻4+,𝑃𝑂43− ) = µ𝑚𝑎𝑥 ( 4
)( 4
)( ) 𝑒( 𝑇
)
𝑋𝐴𝑙𝑔 (Eq.6)
𝐾𝐶𝑂2 +𝑆𝐶𝑂2 𝐾𝑁𝐻+ + 𝑆𝑁𝐻+ 𝐾𝑃𝑂3− + 𝑆𝑃𝑂3− 𝐾2 𝑇+𝐼
4 4 4 4
𝑆𝐶𝑂2 𝑆𝑁𝑂− 𝐾𝑖 𝑁𝐻4+ 𝑆𝑃𝑂3− 𝐾1 𝐼 −𝛽 𝐼

𝜌𝐴𝑙𝑔(𝐶𝑂2,𝑁𝑂3−,𝑃𝑂43−) = µ𝑚𝑎𝑥 3
( )( 4
)( ) 𝑒( 𝑇
)
𝐾𝐶𝑂2 +𝑆𝐶𝑂2 𝐾𝑁𝑂3− +𝑆𝑁𝑂3− 𝐾𝑖 𝑁𝐻4+ +𝑆𝑁𝐻+ 𝐾𝑃𝑂3− + 𝑆𝑃𝑂3− 𝐾2 𝑇+𝐼
4 4 4
𝑆𝐻𝐶𝑂− 𝐾𝑖𝐶𝑂2 𝑆𝑁𝐻+ 𝑆𝑃𝑂3− 𝐾1 𝐼 −𝛽 𝐼

𝜌𝐴𝑙𝑔(𝐻𝐶𝑂3−,𝑁𝐻4+,𝑃𝑂43− ) = µ𝑚𝑎𝑥 3
( 4
)( 4
)( ) 𝑒( 𝑇
)
𝐾𝐻𝐶𝑂−
3
+ 𝑆𝐻𝐶𝑂−
3
𝐾𝑖𝐶𝑂2 + 𝑆𝐶𝑂2 𝐾𝑁𝐻+ + 𝑆𝑁𝐻+ 𝐾𝑃𝑂3− + 𝑆𝑃𝑂3− 𝐾2 𝑇+𝐼
4 4 4 4
𝑆𝐻𝐶𝑂− 𝐾𝑖𝐶𝑂2 𝐾𝑖 𝑁𝐻4+ 𝑆𝑁𝑂− 𝑆𝑃𝑂3− 𝐾1 𝐼 −𝛽 𝐼

𝜌𝐴𝑙𝑔(𝐻𝐶𝑂3−,𝑁𝑂3−,𝑃𝑂43− ) = µ𝑚𝑎𝑥 3
( ) 3
( 4
)( ) 𝑒(𝑇
)
𝑋𝐴𝑙𝑔
𝐾𝐻𝐶𝑂−
3
+ 𝑆𝐻𝐶𝑂−
3
𝐾𝑖𝐶𝑂2 + 𝑆𝐶𝑂2 𝐾𝑖 𝑁𝐻4+ +𝑆𝑁𝐻+ 𝐾𝑁𝑂−
3
+𝑆𝑁𝑂−
3
𝐾𝑃𝑂3− + 𝑆𝑃𝑂3− 𝐾2 𝑇+𝐼
4 4 4
(Eq. 9)
127
Considering the function of light intensity and temperature, parameters 𝐾1 , 𝐾2 and 𝛽 were
fitted to the relative photosynthetic activity (𝑝𝑂2,𝑚𝑎𝑥 ) values for the experiments described in
section 3.1. After minimization of the SSE, values of 𝐾1 = 12.23 (-), 𝐾2 = 21.75 (K-1) and β =
0.013 (-) were obtained. Other biokinetic and physical – chemical parameters were taken from
Decostere et al. (2016).
3.3. Model calibration
Two separate experiments were used for further model calibration. In particular, the maximum
specific growth rate (µmax) and the oxygen mass transfer coefficient (𝐾𝐿 𝑎) were calibrated to
the combined respirometric and titrimetric data. These parameters were considered for model
calibration, because in Decostere et al. (2016), it was illustrated that these parameters are
uniquely identifiable to the combined respirometric and titrimetric data. The first experiment
was performed at a light intensity of I = 10650 lux. The temperature was controlled at 313 K
and the initial microalgal biomass concentration equaled 91.75 g DW m -3. The second
calibration experiment was run at a light intensity of 4810 lux and a temperature of 308 K. The
initial biomass concentration was 104 g DW m-3. Like already mentioned before it was strived
to use similar values of nitrate, phosphate and ammonium for each separate test, respectively
15 g NH4+- N m-3, 0.6 g NO3- - N m-3 and 0.2 g PO43- - P m-3 . In Figure 2, the predicted and
experimental dissolved oxygen profiles and proton addition profiles are illustrated.
For the first calibration experiment a TIC = 0.06 and TIC = 0.05 was calculated for the
respirometric and titrimetric profile, indicating good model performance (Coppens et al., 2014).
Also good visual correspondence between experimental and predicted values was observed.
Considering the second calibration experiment, a visual less good fit (respirometric profile)
was observed. However still low values for TIC were calculated, namely TIC = 0.07 for the
respirometric profile and TIC = 0.05 for the titrimetric profile. The optimized parameter values
for the first experiment were µ𝑚𝑎𝑥 = 0.56 ± 0.0008 d-1 and 𝐾𝐿 𝑎 = 10.02 ± 0.02 d-1. For the
second calibration experiment µ𝑚𝑎𝑥 = 0.62 ± 0.0001 d-1 and 𝐾𝐿 𝑎 = 3.76 ± 0.003 d-1 were
obtained. The optimal values for the maximum specific growth rate were akin for the two
experiments.
3.4. Model validation

One additional experimental runs were used as validation experiment. The mean value of the
optimized values for the maximum specific growth rate was used for the simulations. As such
a value of µ𝑚𝑎𝑥 = 0.59 d-1 was used. Considering the oxygen mass transfer coefficient,
the mean of the two calibration was used. So a value of 𝐾𝐿 𝑎 = 6.43 d-1 was used. The validation
experiment was run with a light intensity of I = 10650 lux and temperature T = 2
128
30
Dissolved oxygen
20
(g O2 m-3)
10
0
0 1 2
Time (d)
Fig 2. Experimental (dashed line) and predicted (full line) dissolved oxygen concentration (left
hand) and proton addition (right hand) for the first (top) and second (bottom) calibration experiment
with Chlorella vulgaris. The first experiment was performed at a light intensity of I = 10650 lux. The
temperature was controlled at 313 K and the initial microalgal biomass concentration equaled
91.75 g DW m-3. The second calibration experiment was run at a light intensity of 4810 lux and a
temperature of 308 K. The initial biomass concentration was 104 g DW m -3.
The initial microalgal biomass concentration was 60.94 g DW m-3. In Figure 3 the predicted
and experimental values of dissolved oxygen (left) and proton addition (right) are depicted.
Good visual correspondence in the first part of the respirometric profile was observed. In the
descending part a slight deviation between experimental and predicted values could be noted.
This could be explained by the difference of experimental and optimized values of oxygen
mass transfer coefficient. Indeed, the experimental value was determined without microalgal
biomass present in the reactor vessel and this can have influence on the parameter (Pittoors
et al., 2014). Considering the titrimetric profile also good correspondence in the first part was
observed and minor deviation between experimental values and predicted values after the
declination point. This can be explained by the fact that the amount of bicarbonate dosed to
the reactor vessel slightly deviated from the foreseen amount. To overcome this drawback, an
additional sensor to online measure the inorganic concentration in the liquid phase might be
suggested. Overall good model performance could be concluded since the calculated TIC
values were 0.19 for the respirometric profile and 0.04 for the titrimetric profile.
129
Fig 3. Experimental (dashed line) and predicted (full line) profiles for the validation
experiments with Chlorella vulgaris. The experiment was run with I= 10650 lux and T= 298.
Initial biomass concentration was 60.94 g DW m -3.
4. CONCLUSIONS
In this research, the combined respirometric and titrimetric methodology was used to assess
the microalgal growth kinetics of a microalgal species that was prior isolated from the biomass
of a WSP. Results showed the interdependent effect of light intensity and temperature on the
growth rate. As such a prior developed model (Decostere et al., 2016) was extended with a
mathematical function which describes this interdependent relationship (Carvalho and
Malcata, 2003). Good model performance with optimized parameter values of maximum
specific growth rate and oxygen mass transfer coefficient was obtained. Further model
validation with an additional experiment illustrated good model performance for both
microalgal species. Next the optimized values for the maximum specific growth rates was
µ𝑚𝑎𝑥 = 0.590 d-1 for Chlorella vulgaris . This value is similar to growth rates reported in
literature.
ACKNOWLEDGMENTS
This research was in the scope of an individual credit for research activities in foreign
countries, with reference number 12Q1617N. This credit was granted by the Foundation for
Scientific Research (FWO), Belgium.
REFERENCES
Alvarado A (2013). Advanced dynamic modelling of wastewater treatment ponds. PhD thesis,
Ghent University, Belgium
APHA (2005). Standard Methods for the Examination of Water and Wastewater (18th edn.),
American Public Health Association, Inc. (APHA) New York, USA
Bordel S, Guieysse B, Muñoz R (2009). Mechanistic model for the reclamation of industrial
wastewaters using algal-bacterial bioreactors. Environmental Sciences Technology,
43, 3200-3207
Box GEP, Draper NR (1987). Empirical Model-building and Response Surfaces. Wiley, New
York, 669
Broekhuizen N, Park JBK, McBride GB, Craggs RJ (2012). Modification, calibration and
verification of the IWA River Water Quality Model to simulate a pilot-scale high rate
algal pond. Water Research, 46, 2911-2926
Carvalho AP, Malcata FX (2003). Kinetic modelling of the autotrophic growth of Pavlova
lutheri: Study of the combined effect of light and temperature. Biotechnology Progress,
19, 1128-1135
130
Coppens J, Decostere B, Van Hulle SWH, Nopens I, Vlaeminck S, De Gelder I and Boon N
(2014). Kinetic exploration of nitrate – accumulating microalgae for nutrient recovery.
Applied Microbiology and Biotechnology, 98, 8377 – 8387
Decostere B, Janssens N, Alvarado A, Maere T, Goethals P, Van Hulle SWH and Nopens I
(2013) A combined respirometer-titrimeter for the determination of microalgae kinetics:
experimental data collection and modelling. Chemical Engineering Journal 222,85–93
Decostere B, De Craene J, Van Hoey S, Vervaeren H, Nopens I, Van Hulle SWH (2016).
Validation of a microalgal growth model accounting with inorganic carbon and nutrient
kinetics for wastewater treatment. Chemical Engineering Journal, 285, 189-197
Haario H, Kalachev L, Laine M (2009). Reduced models of algae growth. Bulletin of
Mathematical Biology, 71, 1626-1648
Henze M, Gujer W, Matsuo T, Van Loosdrecht M (2000). Activated Sludge Models ASM1,
ASM2, ASM2d and ASM3, Scientific and Technical Reports. IWA Publishing, London,
UK
Mannina G, Cosenza A, Viviani G, Vanrolleghem PA, Neumann N (2012). Global sensitivity
analysis for urban water quality modelling : comparison of different methods. 9th
International Conference on Urban Drainage Modelling, Belgrade
Pittoors E, Guo Y and Van Hulle SWH (2014). Modeling Dissolved Oxygen Concentration for
Optimizing Aeration Systems and Reducing Oxygen Consumption in Activated Sludge
Processes: A Review. Chemical Engineering Communications, 201, 983-1002
Richmond A (2004). Chapter 8: Biological principles of mass cultivation. Handbook of
Microalgal Culture: Cornwall, Blackwell Publishing, 125-177
Soetaert K, Herman PMJ (2009). A practical guide to ecological modelling using R as a
simulation platform. Springer-report, New York
Sorokin C, Krauss RW (1962) Effects of temperature and illuminance on Chlorella : Growth
uncoupled from cell division, Plant Physiology, 37, 37–42
Theil H (1961) Economic Forecasts and Policy, 2nd ed., North Holland Publishing Co.,
Amsterdam
Von Sperling M (2007). Waste stabilisation ponds: London, IWA Publishing
Yao B, Xi B, Hu C, Huo S, Su J, Liu H (2011). A model and experimental study of phosphate
uptake kinetics in algae: Considering surface adsorption and P-stress. Journal of
Environmental Sciences, 23, 189 – 198
Yu H, Wilamowski BM (2011). Levenberg–marquardt training. Industrial electronics handbook
5.12, 1-16
131
Aquatic plants and microalgae biomass for the

production of biofuels within a biorefinery
Eugenia J. Olguín*a, Gloria Sánchez-Galvána, Ricardo E. González-Portelaa,
Daniel A. García-Lópeza and Erika Rodríguez-Barradasa
a INSTITUTE OF ECOLOGY (INECOL) Carretera Antigua a Coatepec Km. 2.5,
El Haya, Xalapa 91070 MÉXICO
*Corresponding author: E-mail: eugenia.olguin@inecol.mx Tel. +52 (228) 8421848.
ABSTRACT
Biorefineries are a novel concept for production of biofuels using various types of
biomass feedstocks as replacement of the conventional refinery producing fossil fuels
derived from oil. Waste-water grown aquatic plants and microalgae biomass are a
renewable source of bioenergy and present many advantages over other type of
feedstocks. The aim of the present work was to produce aquatic plant biomass (Pistia
stratiotes) in a phytofiltration lagoon with a potential for biogas generation and the
production of microalgae biomass (Chlorococcum sp.) with a potential for biodiesel
production within a biorefinery at a pilot scale. P. stratiotes was grown in a 13,000 L
phytofiltration lagoon receiving polluted water from an urban river. In the second
production cycle initiating in early January the maximum productivity of P. stratiotes
registered in late February was 11.1 g d.w. m -2 d -1 . Concerning the green microalgae, it
was cultivated in different concentrations of stillage anaerobic effluents (SAE) (10
and 14%) and the highest biomass production (0.6 g L -1 ) was observed using 10%.
Further work was carried out in 18 L flat plate bioreactors using 10% SAE and growth
was higher in these cultures compared to the control synthetic medium. It was
concluded that the P. stratiotes biomass productivity was good enough for harvesting
the biomass in a continuous system for biogas production and that Chlorococcum sp.
is able to grow in a low cost medium containing only stillage anaerobic effluents,
being a promising green microalgae for lipid accumulation.
Keywords: Biofuels, phytoremediation, nutrient recycling, Pistia stratiotes, Chlorococcum sp.
132
Effect of pH, temperature and P/N ratio over a

microalgae consortium for biomass, lipids and
chlorophyll production
Yaremi López Hernández*1, Refugio Rodríguez Vázquez2, Jorge I. Chairez Oria1, Luis
G. Torres Bustillos1,2
1UPIBI-Instituto
Politécnico Nacional. Bioprocesses Department. Acueducto s/n; Barrio La Laguna
Ticoman, 07342; México City. Mexico.
2CINVESTAV. Department of Biotechnology. Av. Instituto Politécnico Nacional 2508, Delegación
Gustavo A. Madero, San Pedro Zacatenco, 07360, México City, Mexico. *Corresponding author:
email: yaremi.lopez.h@gmail.com. Tel 01+(55)57296000 ext 56367.
ABSTRACT
Microalgae are efficient in recovering high amounts of nitrogen, inorganic phosphorus, and
heavy metals from effluents, therefore, have an enormous potential for biotechnological
applications (production of pigments, lipids, foods, and renewable energy). Also wastewater
treatment is an interesting option. Microalgal biomass can generate bioenergy in a number of
ways. 1) By triacyl-glicerides production which can be esterified to produce biodiesel, 2) after
the extraction of lipids, proteins and pigments of high added value, biomass itself can be
hydrolyzed to produce more sugars and then by means of yeast, convert them into bioetanol,
3) the residual biomass can be digested anaerobically to produce bio-methane. This work had
two aims: first was evaluate a microalgae consortium (composed by Chlorella, Scenedesmus
and some diatoms, including Nitzchia) regarding their capabilities of N and P removal from a
defined medium resembling a real wastewater. A second aim was to characterize the biomass,
lipids and chlorophyll produced by consortia varying four parameters at three levels. Factors
i.e., temperature (25-35 °C), pH (8-10), N and P (N/P ratios between 4 and 40) were varied in
accord to Taguchi`s fractional factorial design -4 factors at three Levels-. Assessments were
carried out during 21 days under agitation (140 rpm) and irradiance of 100 μE/m 2s. Biomass
as dry biomass, N and P removals, lipids and chlorophyll a and b were evaluated. The
parameter which affected mostly to biomass, lipids and chlorophyll production was
temperature (30 °C). pH and N/P ratio also resulted to be important. Best biomass production
was found at 30°C, pH 9 and N/P = 40 ratio. A maximum biomass concentration of 1.3 g/L
was obtained though the production of lipids and sum of chlorophyll (a+b) was the second
highest value found in this work: 35 and 2.55 mg L-1, respectively. The maximum lipid
production was observed for the same temperature, but with a pH of 8 and N/P = 13. This
value was of 40 mg L-1, with a total biomass and chlorophyll values of 1.21 and 1.48 mg L-1,
respectively. Finally, the maximum chlorophyll production was achieved at 30oC, with a pH of
10 and N/P ratio = 4. In conclusion, using the microalgae consortia, it was feasible to reach
maximum concentrations of biomass, lipids and total chlorophyll of 1,300, 40 and 54 mg L-1,
with productivities of 61.6, 1.89 and 2.57 mg L-1d-1 respectively. The parameter which affected
more the process was temperature followed by pH and the ratio N/P at the end.
Keywords: Chlorella, lipids, N/P removal, wastewaters
133
1. INTRODUCTION
Microalgae are efficient in recovering high amounts of nitrogen, inorganic phosphorus, and
heavy metals from effluents, therefore, have an enormous potential for biotechnological
applications (production of pigments, lipids, foods, and renewable energy). Also wastewater
treatment is an interesting option. Microalgal biomass can generate bioenergy in a number of
ways. 1) By triacyl-glicerides production which can be esterified to produce biodiesel, 2) after
the extraction of lipids, proteins and pigments of high added value, biomass itself can be
hydrolyzed to produce more sugars and then by means of yeast, convert them into bioetanol,
3) the residual biomass can be digested anaerobically to produce bio-methane. Many authors
have reported the effect of parameters such as pH, temperature and nutrients content over
growth and lipid production. Butterwick et. al. (2005) studied the effect of temperature in the
diversity and growth rates of freshwater algae. They studied the effect of temperature (2-35oC)
over the growth rate of 21 freshwater microalgae. They found that some strains were capable
of growing at extreme temperatures of 2 and 35oC, but most of the strains grew well in the 7-
25oC interval. Strains capable of growth at 35oC was Aphanizomenon flos-aquae, while those
which survive and growth at 30°C were Eudorina unicocca, Staustratum cingulum, Limnothrix
redekei, Anaeba flos-aquae, Microsistis aueruginosa and Fragilaria cotonensis.
Rai et. al.(2014), studied the effect of pH, salinity and temperature on the growth of six species
of marine phytoplankton. These authors concluded that the isolates preferred near to neutral
to alkaline pH. Growth was adversely affected over a pH= 10.6. All isolates showed best
growth at 20 and 30oC and growth ceased at 40-50oC. Regarding salinity, all isolated grew
well at all salinity levels assessed, below 40 mg/L. Nowrozi and Valavi (2011) studied the
effects of environmental factors on phytoplankton abundance and diversity. They concluded
that temperature was a determinant factor for phytoplankton growth and diversity, but other
factor such as pH and nutrient contents influenced the diversity indexes. On the other hand,
the works of Beuckels et. al. (2015) are very clarifying regarding the effect of N and P
concentration over the removal of those nutrients by microalgae in wastewaters. They
demonstrated that microalgae can adjust the N and P concentrations in their biomass to the
N and P supply in the wastewater. Finally they concluded that sufficiently high N
concentrations are needed to ensure effective P removal from wastewater due to the positive
effect of N on the accumulation of P.
Finally, Bartley et al (2014) reported the effect on growth and lipid accumulation of the
microalgae Nannochloropsis salina. They observed that this microalgae was susceptible to
pH changes: N. salina showed highest growth rates at pH 8 and 9. Lipid accumulation of
unbuffered controls produced 21% of FAME but no additional lipid content was observed by
manipulating pH values at the beginning of the stationary phase. This work had two aims: first
was evaluate a microalgae consortium (composed by Chlorella, Scenedesmus and some
diatoms, including Nitzchia) regarding their capabilities of N and P removal from a defined
medium resembling a real wastewater. A second aim was to characterize the biomass, lipids
and chlorophyll produced by consortia varying the four parameters (ph, temperature, N and P
contents) at three levels.

The employed consortium is composed of Chlorella vulgaris, Scenedesmus sp. and some
diatoms, including Nitzchia. It was kindly donated by UMB-Unidad Tultitlan (Edo. de Mexico,
México). The strain was conserved by successive inoculation of BG-11 medium with 10%
inoculum every 2-3 weeks.
134
Assessments were carried out in 500 mL baffled flask, containing 300 mL of BG-11 medium
conditioned at different pH values (i.e. 8, 9 and 10) and N-NO3 (250, 625 and 1,000 mg/L)and
P contents as PO4- (25, 50 and 70 mg/L),N/P ratios from 4 to 40. Experiments were carried
out in a closed agitation system were flasks were agitated at 140 rpm during 21 days,
illuminated using LED stripes with a total average irradiance of 100 E/m2s.Temperatures
were controlled at 25, 30 and 35oC. An experimental design using a modified ortogonal
Taguchi experimental design (Amaral, et.al. 2015) it is shown at Table 1. All assessments
were carried out by triplicate.
During the process, biomass concentrations were measured by dry weight and absorbance,
N-NO3 and P-PO4 were monitoring by colorimetric reactions with diphenyl-sulfonic acid and
molivdovanate method, respectively. At the end of the process, lipids, chlorophyll a and b were
extracted by organic solvents and contents were measured. Biomass, lipid and chlorophyll
(a+b) productivities were calculated as well as the P and N removal rates.
Table 1. Experimental design Taguchi (L9 –A 34-2 Fractional Factorial Design 4 Factors at
Three Levels).
o
Flask pH Temperature ( C) Nitrogen (mg/L) Phosphorus (mg/L) N/P ratio
1 8 25 250 25 10
2 8 30 625 50 13
3 8 35 1,000 70 14
4 9 25 625 70 9
5 9 30 1,000 25 40
6 9 35 250 50 5
7 10 25 1,000 50 20
8 10 30 250 70 4
9 10 35 625 25 25

After the assessments, the color and appearance of the flasks containing the microalga
cultures were very different. Specifically the temperature of the process seems to affect the
composition of the consortium. As an example, in figures 1 a, b and c, the final view of the
culture flasks are shown.
Figure 1. Experiments carried out at: a) 25oC; b) 30oC and c) 35oC. Culture flasks after 21
days of growth.
At 25oC (Figure 1a), the culture presented coagulation and sedimentation. Color was dry green
and some adherence to the flask walls was observed. Microscopy showed domination of
Scenedesmus in 4 and 6 individuals groups. On the other hand, at 30oC (Figure 1b) no
coagulation/sedimentation was observed. The microalgae showed adherence to the flask
135
walls, but in a lower extent than that observed at 25oC. Microscopy showed the existence of
a filamentous microalgae and Scenedesmus in a greater extent. The color was emerald-green.
Finally, at 35oC (Figure 1c), the culture turned light green, there was some agglomeration and
the microscopy showed prevalence of the filamentous microalgae. These simple observations
suggest that temperature played an important role in the culture of the consortium. At the time,
the identification of the consortium components using molecular techniques is on process. In
that way, a precise description of the genera and species present and how the environmental
parameters (temperature, pH, nutrients) affect the constitution of the consortium will be
established.
Main results of the assessments regarding biomass, lipids and chlorophyll contents are
summarized on Table 2. It is interesting to observe that the higher biomass concentrations
were achieved at 30oC. The higher value (1.29 g L-1) was obtained with conditions prevailing
in flask 5. Regarding the higher lipid production it was found in flask 2 (40 mg L-1). Finally, the
higher chlorophyll (a+b) value corresponded to flask 8 (54.06 mg/L). Note that these
assessments (2, 5 and 8) were carried out at 30oC). All these values can be expressed in
terms of productivities, giving PX, PL and PCh of 61.6, 1.9 and 2.6 mgL-1d-1 for biomass, lipids
and chlorophyll, respectively. It is remarkable that the different culture conditions had different
effects over the biomass, lipids and chlorophyll final concentrations. In other way: there is no
condition where maximum biomass, lipids and chlorophyll concentrations are optimum. There
exists a compromise among the three products. Nevertheless, there are some conditions
where fair production of biomass lipids and chlorophyll (a+b) can be achieved, such as flask
5 (pH=9, 30oC, N= 1,000 mgL-1 and P= 25 mgL-1, N/P = 40), where 1,290, 35 and 53.4 mgL-1
of every product were obtained, respectively.
N and P removals for the different conditions resulted as follows (See Table 3). Maximum N
removals were obtained for assessments 1, 6 and 8 (those with an initial N = 250 mg L-1).
Table 2. Consolidated results for the nine assessments (average of triplicates)

Flask Xmax Lipids Chlorophyll PX PL PCh
(gL-1) (mgL-1) (a+b) (mgL-1) mgL-1d-1) mgL-1d-1) (mgL-1d-1)
1 0.8633 2 26.20 41.11 0.1052 1.24
2 1.21 40 31.17 57.77 1.88 1.48
3 0.9000 24 9.25 42.85 1.14 0.4404
4 0.8537 1.4 43.00 40.65 0.0666 2.04
5 1.29 35 53.47 61.58 1.65 2.54
6 0.8700 17 6.10 41.42 0.8201 0.2903
7 1.16 2.0 44,73 55.23 0.0947 2.12
8 1.15 26 54.06 54.92 1.23 2.57
9 0.9400 21 6.51 44.76 1.02 0.3098
Nearly 100% of N removal was observed. On the other hand, the maximum P removals were
achieved for assessments 1,5 and 9 (where P = 25 mg L-1). In general it is obvious that N and
P removals resulted inversely proportional to the initial N and P concentrations. If the N and P
removal rates are calculated, the view is quite different. Maximum N (28.57 mg L -1d-1) and P
(0.8326 mg L-1d-1) removal rates were found for assessments5 and 1. In contrast, minimum N
(10.73 mgL-1d-1) and P (0.2785 mgL-1d-1) removal rates were found for assessments 3 and 7,
respectively.
136
Table 3. N and P removals and removal rates (average of triplicates).

Flask N removal N removal rate P removal P removal rate
(%) (mgL-1d-1) (%) (mg L-1d-1)
1 100.0 11.90 69.94 0.8326
2 58.98 17.55 32.18 0.7664
3 22.55 10.73 9.90 0.3333
4 53.14 15.81 11.15 0.3716
5 60.01 28.57 63.42 0.7550
6 100.0 11.90 33.10 0.7880
7 30.68 14.60 11.7 0.2785
8 98.67 11.74 17.49 0.5830
9 44.26 13.17 59.45 0.7077
4. CONCLUSIONS
The visual observations of the culture flasks at the end of growth assessments at 20, 25 and
30oC suggest that temperature played an important role in the culture of the consortium. At
the time, the identification of the consortium components is carried out, using molecular
techniques. In that way, a precise description of the genera and species present and how the
environmental parameters (temperature, pH, nutrients) affect the constitution of the
consortium, will be established. The parameter which affected mostly to biomass, lipids and
chlorophyll production was temperature (30 °C). pH and N/P ratio also resulted to be
important. Best biomass production was found at 30°C, pH 9 and N/P = 40 ratio. A maximum
biomass concentration of1.3 gL-1 was obtained though the production of lipids and sum of
chlorophyll (a+b) was the second highest value found in this work: 35 mgL-1 and 2.55 mgL-1,
respectively. The maximum lipid production was observed for the same temperature, but with
a pH of 8 and N/P = 13. This value was of 40 mg/L, with a total biomass and chlorophyll values
of 1.21 and 1.48 mgL-1, respectively. Finally, the maximum chlorophyll production was
achieved at 30oC, with a pH of 10 and N/P ratio = 4. In conclusion, using the microalgae
consortia, it was feasible to reach maximum concentrations of biomass, lipids and total
chlorophyll of 1.3, 40 and 54 mgL-1, with productivities of 61.6, 1.89 and 2.57 mgL-1d-1,
respectively. The parameter which affected more the process was temperature followed by
pH and the ratio N/P at the end.
ACKNOWLEDGMENTS
The financial support of SIP-IPN project is acknowledged and thanked.
REFERENCES
Amaral MS, Loures CC, Da Ros PCM, Machado SA, Reis CER, de Castro HF, Solva MB
(2015) Evaluation of the cultivation of marine microalgae Chlorella sp. to be used as
feedstock in ultrasound assisted ethanolysis. Biofuel Research Journal 7: 288-294.
Bartley ML, Bocing WJ, Dungan BN, Holguin FO, Schaub T (2014) pH effects on growth and
lipid accumulation of the biodiesel microalgae Nannocloropsis salina and invading
organisms. Journal of Applied Phycology. 26: 1431-1437.
Beuckels A, Smolders E, Muylaert K (2015) Nitrogen availability influence phosphorus
removal in microalgae-based wastewater treatment. Water Research. 77: 98-106.
Butterwick C, Heaney SI and Talling JF (2005) Diversity in the influence of temperature on
growth rates of freshwater algae, and its ecological relevance. Freshwater Biology. 50:
291-300
137
Nowrouzi SandValavi H (2011) Effects of environmental factors on phytoplankton abundance

and diversity in Kafar Lake. Journal of Fisheries and Aquatic Science. 6(2): 130-140.
Rai SV and Rajashekhar M (2014) Effect of pH, salinity and temperature on the growth of six
species of marine phytoplankton. Journal of Algal Biomass Utilization. 5(4): 55-59.
138
Production of chlorella vulgaris in culture media

with different source of nitrogen and carbon to
obtain lipids
Citlally Ramírez-López*, Jorge Isaac Chairez-Oria, Luis Carlos Fernández-Linares*
Departamento de Bioprocesos, Unidad Profesional Interdisciplinaria de Biotecnología,

Instituto Politécnico Nacional, México.
*Corresponding author: citlallyrl@yahoo.com.mx, lfernand36@hotmail.com. Tel. +52 57296000 Ext.
56387
ABSTRACT
An optimal medium formulation is critical to guarantee sufficient and stable supply of nutrients
to ensure a maximal microalgal growth rate, and to produce biofuels at higher efficiencies. A
new medium named M was designed for Chlorella vulgaris UTEX 26, considering the
composition of the BBM (Bold's Basal Medium) and HAMGM (Highly Assimilable Minimal
Growth Medium) culture media. To check that M medium could enhance biomass and lipid
productions higher than conventional culture media, C. vulgaris UTEX 26 was grown with M
medium and three additional modified culture media: pH-M (M medium to constant pH), M-U
(NH4HCO3 was replaced with urea) and M-UBS (NH4HCO3 was replaced with urea and
NaHCO3). Results of these culture media were compared with the BBM and HAMGM culture
media. The highest biomass concentration (0.87 g L-1) was observed with M-UBS medium,
which did no differ significantly of M medium. This biomass concentration was 40% higher
than BBM medium. On the other hand, the M-U medium barely supported microalgal growth.
Regarding lipid concentration, the highest lipid concentration was produced with M medium
(281 mg L-1), which it was 70, 75 and 85% higher than the BBM, M-UBS and HAMGM culture
media, respectively. Fatty acids (C16:0, C18:0, C18:1, C18:2 and C18:3) produced by C.
vulgaris UTEX 26 are suitable for producing biodiesel. M medium allowed rising the biomass
and lipid concentrations of C. vulgaris UTEX 26, on 20 days of culture.
Keywords: Chlorella vulgaris; ammonium bicarbonate; urea; fatty acids.
1. INTRODUCTION
Microalgae produce a great variety of compounds. In recent years, the biodiesel production
from its lipids has emerged as a good alternative (Giordano et al., 2014). The culture
conditions that promote the production of biomass and lipids at low cost are central to the
research and development of biodiesel production. The optimal formulation of culture medium
is essential to ensure an adequate supply of nutrients that promotes maximum growth rate,
high productivity of biomass, lipids and metabolites of interest (Schenk et al., 2008). To
increase the productivity of biomass and lipids is necessary that the bicarbonate ion or CO 2
are added to the medium (Kee Lam et al., 2013). Similarly, the source and concentrations of
nitrogen (Hadj-Romdhane et al., 2012) and phosphorus (Chu et al., 2013) and nitrogen
limitation (Klok et al., 2013) are crucial to the productivity of biomass and lipids.
139
Conventional culture media are not specific to a particular microalgae. The Bold’s Basal
Medium (BBM) is widely used for various kinds of microalgae (Blair et al., 2014). Meanwhile,
the Highly Assimilable Minimal Growth Medium (HAMGM) was designed by Hadj-Romdhane
et al. (2012) for Chlorella vulgaris culture in an air-lift photobioreactor (PBR) in continuous
regimen and with the objective of recycling the wasted culture medium. Microalgal growth in
raceway, exhibits low biomass productivity compared with the obtained in an air-lift PBR (Fon
Sing et al., 2013), thereby, the nutrients of HAMGM medium could be in excess when the
culture is performed in raceway. The main differences between BBM and HAMGM culture
media are the source and amount of nitrogen and carbon. BBM medium contains 0.25 g L-1 of
NaNO3 as nitrogen source, while HAMGM medium contains 1.099 g L-1 of NH4HCO3 as
nitrogen and carbon source.
In the present study, the nitrogen and carbon source of M medium were modified to determine
the effect on the production of biomass and lipids and compared with both, BBM and HAMGM
culture media.

A new culture medium named M was designed for the growth of Chlorella vulgaris UTEX 26,
considering the composition of the BBM (Bold's Basal Medium) and HAMGM (Highly
Assimilable Minimal Growth Medium) and a strategy of three stages. To check that M medium
could enhance biomass and lipid productions higher than conventional culture media, C.
vulgaris UTEX 26 was grown with M medium and three additional modified culture media: pH-
M (M medium to constant pH), M-U (NH4HCO3 was replaced with urea) and M-UBS (NH4HCO3
was replaced with urea and NaHCO3). Results of these culture media were compared with the
BBM and HAMGM culture media. The response variables were the concentrations of biomass
and lipids. Biomass concentration was determined at 0, 4, 8, 12, 14, 16 and 20 days of culture.
Lipid concentration was determined on days 16 and 20. The experiments were performed in
a 1 L PBR with a working volume of 0.9 L, and bubbled with air at a constant flow rate of 0.5
vvm. The PBR was illuminated with light intensity of 180 μmol photons m -2s-1 on the surface
of the reactor by using cool white fluorescent tubes. Dry biomass was determined
gravimetrically. Lipids were quantified gravimetrically using hexane and isopropanol for the
extraction after an acid hydrolysis of biomass. Lipid composition was determined as FAMES
by gas chromatography after lipid transesterification (Ramírez-López et al., 2016). Moreover,
consumption of NH4+ (APHA-AWA-WPCF, 1992) and NO3- (modified of Keeney and Nelson,
1982) was quantified. Analyses were performed in triplicate and the results are given as the
average of three measurements. The results were subjected to one way ANOVA at a 95%
confidence limit (p>0.05).
The results of biomass and lipid production from C. vulgaris UTEX 26 achieved with the M
medium were compared with the pH-M, M-U, M-UBS, BBM and HAMGM culture media. For
our culture conditions, M medium does not contain an excess of nitrogen as HAMGM medium,
and if required, the M medium could be recycled because it does not contain any poorly
assimilated ions (Na+ or Cl-) by C. vulgaris (Hadj-Romdhane et al., 2012).
The M and M-UBS culture media performed a positive effect on C. vulgaris UTEX 26 growth
(Fig. 1). Since 8 days of culture had significant differences (p>0.05) between the culture media,
because biomass concentration with M y M-UBS culture media were threefold higher than in
the M-U and BBM culture media. Significant differences (p<0.05) between biomass
concentrations obtained with the M and BBM culture media continued until the end of the
microalgal culture. On the other hand, from the twelfth day until the end of the microalgal
culture, no differences between M and HAMGM culture media were observed. On day 16, all
140
the available sources of nitrogen of the culture media had been consumed, except for pH-M
and HAMGM culture media. C. vulgaris UTEX 26 cultivated in M medium exhibited a μ of
0.23d-1 and a biomass concentration of 0.66 g L-1; which was 61% higher than the biomass
obtained with BBM medium and threefold higher than the biomass concentration of C.
sorokiniana cultured by Ramanna et al. (2014), in wastewater with a nitrogen concentration
ninefold higher than the included in the M medium.
Figure 21. Profiles of biomass concentrations, as a function of culture time,

for Chlorella vulgaris UTEX 26 in different culture media.
The highest biomass concentration was observed at 20 days of culture with M-UBS medium
(0.87 g L-1), which did no differ significantly of M medium and these results were 45% higher
than biomass concentration of C. vulgaris obtained by Yeh et al. (2010) with 152% more
carbon source. Moreover, NH4+ and urea are used for cultivation of microalgae due to the
relatively low cost. The NH4+ is directly absorbed by passive diffusion and urea dissociates to
HCO3- and NH4+ in solution via the urea amidohydrolase pathway (urease enzyme). This NH4+
is accumulates to form amino acids. It was also found that biomass yields and lipid production
increased with CO2 addition. The additional CO2 dissociated may have a dual effect by
augmenting microalgal growth as well as providing excess carbon flux towards lipid production
(Moheimani, 2013). These could be the reasons by which urea with NaHCO 3 and NH4HCO3
(M-UBS and M culture media) show to be the most effective nitrogen and carbon sources to
obtain the highest biomass concentration.
Furthermore, it has been reported by Mandalam and Palsson (1998) and Blair et al. (2014)
that addition of major elements (nitrogen, phosphorus or magnesium) can increase the overall
capacity of a culture medium to produce biomass, however, with the culture conditions used
in this work the result were different. The HAMGM medium contains a nitrogen and carbon
source 146% higher than the M medium and there was no significant difference for biomass
production between the M and HAMGM culture media. In contradistinction to M and M-UBS
media, the M-U medium with urea as nitrogen and carbon source, it barely supported
microalgal growth due to the pH was decreasing after day 4 until an average pH of 3.7 after 8
days of culture, and acidic pH values cause enzyme inactivation (Kumar and Das, 2012).
Similar results were reported by Ramanna et al. (2014) with C. sorokiniana and NH4NO3 as
nitrogen source. In the M y HAMGM culture media, when more than 90% of available NH 4+
had been consumed, the pH decreased until 4. On the other hand, the M-UBS and BBM
culture media were maintained an alkaline pH (between 7.5 y 10.0), which it was favorable for
the C. vulgaris UTEX 26 growth.
141
Regarding lipid concentration, there was no significant difference among the productions
obtained on day 16 with the culture media, except with pH-M medium (Fig. 2). At the end of
the culture, the highest lipid concentration was obtained with M medium (281 mg L-1), which
was significantly different (p<0.05) of the other culture media. The increase of lipid production
in M medium could be due to the four days of nitrogen starvation after 16 days of culture.
Similar results were reported by Widjaja et al. (2009), when they reported about factors
influencing lipid production on C. vulgaris and their results suggested that lipid concentration
was higher after seven days with nitrogen starvation. Different results in lipid concentration
between M and pH-M media could be due to a stress by decreasing of pH in M medium.
Figure 22. Lipid concentrations for Chlorella vulgaris UTEX

26 in different culture media on days 16 and 20 of culture.
Gas chromatograph analysis of lipids extracted from C. vulgaris UTEX 26 grown on M, pH-M,
M-U, M-UBS, HAMGM and BBM culture media showed a fatty acid profile that consisted of
fatty acids with carbon chains ranging from C15:0 to C18:0 on days 16 and 20 of culture. The
majority of the fatty acids were C16:0 (palmitic acid), C18:0 (stearic acid), C18:1 (oleic acid),
C18:2 (linoleic acid) and C18:3 (linolenic acid), which are more suitable for obtaining ideal
biodiesel properties (Cho et al., 2015; Jusoh et al., 2015). Ramanna et al. (2014) found that
C. sorokiniana grown in wastewater also showed a similar fatty acid profile with the majority
of fatty acids comprising of carbon chain lengths of C14–C18.
4. CONCLUSIONS
The present study demonstrated that the M medium for C. vulgaris UTEX 26 improved
biomass and lipid concentrations (40% and 85%, respectively) with respect to the conventional
culture media. Fatty acids (C16:0, C18:0, C18:1, C18:2 and C18:3) produced by C. vulgaris
UTEX 26 are suitable for producing biodiesel. The M and M-UBS culture media allowed to
decrease until 50% some nutrient concentrations required in the conventional culture media,
moreover, medium recycling could be also possible.
142
ACKNOWLEDGMENTS
The authors are grateful for the financial support provided by the Unidad Profesional
Interdisciplinaria de Biotecnología of Instituto Politécnico Nacional (grant SIP 20161677), to
SENER (Secretaría de Energía, grant 174627) and to CONACyT (Consejo Nacional de
Ciencia y Tecnología, grant 233384).
REFERENCES
APHA-AWA-WPCF (1992) Standard Methods for the examination of water and wastewater. 18th Edition.
American public Health Association. Washington, D. C. U.S.A.
Blair M F, Kokabian B, Gude V G (2014) Light and growth medium effect on Chlorella vulgaris biomass
production. J Environ Chem Eng 2:665-674.
Cho H U, Kim Y M, Choi Y-N, Xu X, Shin D Y, Park J M (2015) Effects of pH control and concentration
on microbial oil production from Chlorella vulgaris cultivated in the effluent of a low-cost organic
waste fermentation system producing volatile fatty acids. Bioresour Technol 184:245-250.
Chu F F, Chu P N, Cai P J, Li W W, Lam P K S, Zeng R J (2013) Phosphorus plays an important role
in enhancing biodiesel productivity of Chlorella vulgaris under nitrogen deficiency. Bioresour
Technol 134:341-346.
Fon Sing S, Isdepsky A, Borowitzka M A, Moheimani N R (2013) Production of biofuels from microalgae.
Mitig Adapt Strateg Glob Change 18:47-72.
Giordano P C, Beccaria, A J, Goicoechea H. C (2014) Rational design of a culture medium for the
intensification of lipid storage in Chlorella sp. Performance evaluation in air-lift bioreactor.
Bioenergy Research 158:269-277.
Hadj-Romdhane F, Jaouen P, Pruvost J, Grizeau D, Van Vooren, G, Bourseau P (2012) Development
and validation of a minimal growth medium for recycling Chlorella vulgaris culture. Bioresour
Technol 123:366-374.
Hu Q, Sommerfeld M, Jarvis E, Ghirardi M, Posewitz M, Seibert M, Darzins A (2008) Microalgal
triacylglycerols as feedstocks for biofuel production: perspectives and advances. Plant. J.
54:621-639.
Jusoh M, Loh S H, Chuah T S, Aziz A, Cha T S (2015) Elucidating the role of jasmonic acid in oil
accumulation, fatty acid composition and gene expression in Chlorella vulgaris
(Trebouxiophyceae) Turing early stationary growth phase. Algal Res. 9:14-20.
Kee Lam M, Teong Lee K, Rahman-Mohamed A (2013) Current status and challenges on microalgae-
based carbon capture. International Journal of Greenhouse Gas Control. 10:456-469.
Keeney D R, Nelson D W (1982) Nitrogen Inorganic Forms. In: Black C A (Ed) Methods of soil analysis
Part 2: Agronomy. American Society of Agronomy. Madison, Wisconsin, U.S.A. pp 643-698.
Klok A J, Martens D E, Wijffels R H, Lamers P (2013). Simultaneous growth and neutral lipid
accumulation in microalgae. Bioresour Technol 134:233-243.
Kumar K, Das D (2012) Growth characteristics of Chlorella sorokiniana in airlift and bubble column
photobioreactors. Bioresour. Technol. 116:307-313.
Liu N, Li F, Ge F, Tao N, Zhou Q, Wong M (2015) Mechanisms of ammonium assimilation by Chlorella
vulgaris F1068: Isotope fractionation and proteomic approaches. Bioresour Technol 190:307-
314.
Mandalam R K, Palsson B Ø (1998) Elemental balancing of biomass and medium composition
enhances growth capacity in high-density Chlorella vulgaris cultures. Biotechnol Bioeng
59:605-611.
Moheimani N R (2013) Inorganic carbon and pH effect on growth and lipid productivity of Tetraselmis
suecica and Chlorella sp (Chlorophyta) grown outdoors in bag photobioreactors. J. Appl.
Phycol. 25:387-398.
Ramanna L, Guldhe A, Rawat I, Bux F (2014) The optimization of biomass and lipid yields of Chlorella
sorokiniana when using wastewater supplemented with different nitrogen sources. Bioresour
Technol 168:127-135.
Ramírez-López C, Chairez I, Fernández-Linares L (2016) A novel culture medium designed for the
simultaneous enhancement of biomass and lipid production by Chlorella vulgaris UTEX 26.
Bioresour Technol 212:207-216.
Schenk P M, Thomas-Hall S R, Stephens E, Marx U C, Mussgnug J H, Posten C, Kruse O, Hankamer
B (2008) Second Generation Biofuels: High-Efficiency for Biodiesel Production. Bioenergy
Research. 1:20-43.
Widjaja A, Chien C-C, Ju Y-H (2009) Study of increasing lipid production from fresh water microalgae
Chlorella vulgaris. J. Taiwan Inst. Chem. Eng. 40:13-20.
143
Xin L, Hong-ying H, Ke G, Ying-xue S (2010) Effects of different nitrogen and phosphorus

concentrations on the growth, nutrient uptake, and lipid accumulation of a freshwater microalga
Scenedesmus sp. Bioresour Technol 101:5494-5500.
Yeh K, Chang J, Chen W (2010) Effect of light supply and carbon source on cell growth and cellular
composition of a newly isolated microalga Chlorella vulgaris ESP-31. Engineering in Life
Sciences 10:201-208.
144
Oral Presentations – Residues Valorization II

145
Availability of agro residues and estimation of the

theoretical, sustainable and technical energetic
potentials in Mexico: The scenario in the last ten
years
Jessica Denisse Valle Garcíaa, Danahé Marmolejo Correab and Edgar Vázquez Núñeza*
a Department of Chemical, Electronics and Biomedical Engineering, University of Guanajuato. Lomas

del Bosque 103, León, Guanajuato, México, MX37150. a E-mail: vallegj2013@licifug.ugto.mx; a
Corresponding author email: edgar.vazquez@ugto.mx
b Department of Physics Engineering, University of Guanajuato, Lomas del Bosque 103, León,
Guanajuato, México, MX37150. b E-mail: marmolejocorrea@ugto.mx
ABSTRACT
The use of biomass as a renewable energy source minimizes the negative environmental
impacts, produces minimum secondary wastes after its use and is sustainable in both,
economic and social scenarios. The amount of agriculture residues in Mexico is considerable
large. Close to 55% of the soil in Mexico is destined to agriculture purposes. Thus, the use of
these residues as input for power production becomes very attractive. In Mexico, the most
important cultivars are corn, wheat, sorghum and barley. These were listed according to their
total cultivated area and its economic value. This work focuses on the availability and
localization analyses of the agro residues generated in Mexico and on the estimation of its
energetic potential. The study shows statistic values of the agro residues from the last ten
years in the 32 Mexican states. The theoretical, technical and sustainable potentials were
calculated, considering physical and chemical parameters of the analyzed residue, i.e.,
Residue to Product Ratio (RPR), humidity, Lower and Higher Heating Values (LHV & HHV).
In the calculation of the Theoretical Potential (TheP) all residues were contemplated. The
sustainable potential (SuP) includes the conservation of organic matter to avoid negative
effects on soil. The technical potential (TechP) is defined as the fraction of the theoretical
potential minus the organic matter destined for competitive or multiple uses i.e., feeding of
livestock or human construction. This work concludes that by using data from the last ten
years, it was possible to calculate the Energy Potential using a more sustainable approach.
Keywords: Mexico; Energetic Potential; Sustainability; Agricultural Residues
1. INTRODUCTION
Biomass refers to solid carbonaceous material derived from plants and animals (Jhonson et
al., 2007). These include the residues of agriculture and forestry, animal wastes and wastes
from food processing operations. The biomass is generated predominantly in rural areas, it
also provides an important fuel source for the urban poor and many rural, small and medium
scale industries. This production of biomass in all its forms for fuel, food and fodder demands
environmentally sustainable land use and integrated planning approaches (Scarlat and
Dallemand, 2011). Biomass is generally and erroneously regarded as a low status fuel, and
rarely finds its way into the energy statistics when, in fact, it should be considered as are
newable energy equivalent to fossil fuels. It offers considerable flexibility of fuel supply due to
the range and diversity of fuels that can be produced. Biomass can be burnt directly or it can
146
be converted into solid, gaseous and liquid fuels using conversion technologies such as
fermentation to produce alcohols, bacterial digestion to produce biogas and gasification to
produce a natural gas substitute. Industrial, agricultural livestock and forest residues can be
used as a biomass energy source, or energy crops such as trees and sugarcane can be grown
specifically for conversión to energy. In developing countries, traditional biomass fuels
contribute to the largest share of the total energy uses where up to 95% of national energy
consumption relies on biomass. Whereas in developed countries, biomass is converted into
cleaner and efficient energy i.e. electricity. In developed countries, an efficient biomass use is
becoming important agenda for reducing GHGs and generating renewable energy (Gurung
and Oh, 2013). The traditional biomass sources in developing and least developed countries
can be converted into ‘efficient bioenergy’ in the forms of gas, liquid and electricity.
1.1. The Mexican scenario
Mexico is a Latin American country with an anual GDP grothw of 3.2% and a Gross National
Income (GNI) of USD 9,870 per person per year (WorldBank, 2014). The private consumption
has been the most important driver during the last five years, also the investments on the
industrial and communication sectors has given support to the national economy (Alemán-
Nava et al., 2014). Mexico has a predominantly service dependency economy with the 62.3%
of the GDP, the industrial sector is valued as the 34.4% of the GDP. The agriculture represents
the 3.3 % of the GDP, however besides the economical value this sector has, the energy
supply potential by using the agro residues remains not enterily exploited. The use of biomass
as a primary source of energy has been decreasing in Mexico since 1965, when it constituted
15.3% of the total primary energy supply. As of year 2005, this share represented only 5.3%.
Meanwhile, the use of hydrocarbon fuels has been steadily increasing and accounted for
87.7% of the gross primary energy supply (Alemán-Nava et al., 2014). So far, there are
reasons for researching about the use of biomass as energy resource in a country which has
petroleum:
a) Global problematic related to the fossil fuels in the globe.

b) Reduction of the national oil reserves, generating unestability in the national and global
economies.
c) The generation of GHG from the fossil fuels.
In contrast, the use of biomass as energy source has the possibility to be a sustainable energy
source, impacting not only in the Mexican energy diversification, but also in the generation of
new energy technologies (Islas et al., 2007). Some studies have suggested theoretical values
for the Mexican scenario, Ghilardi et al. (2005) proposes the values between 3035 PJ/year
and 4550 PJ/year.
1.2 Biomass resources in Mexico
Mexico has an high electrification rate, the 99.1% of the population have acces to electricity
(WorldBank, 2014), however, according to the global environmental politics the use of biomass
generated in agricultura could be pivotal for reducing the Green House Gas emissions (GHG)
and impacting on the global environmental politics. Mexico has enormous potential of energy
from renewable sources including not only the biomass source, but solar and wind (Ramírez
et al., 2010). Some estimations have been done considering the most important primary and
secundary residues, among the most relevant are: barley, chickpea, lentil, wheat, canola,
cotton, sugarcane, etc. (Váldez-Vázquez et al., 2010). In this way, Mexico coulb become an
important producer – consumer of biomass for producin bioenergy.
147
1.3 Technologies for biomass exploitation
There are different technological pathways from which the biofuels can be produced from the
lignocellulosic biomass. These technologies include thermochemical or the biochemical
processing pathways (Cherubini et al., 2010). The thermochemical processes are combustion,
gasification, pyrolysis, liquefaction, hydrothermal and biochemical conversión including
fermentation and anaerobic digestion. The various thermal treatments can be broadly
classified into combustion, gasification and pyrolysis.
a. Combustion process involves the direct burning of the biomass material in air. This method
of biomass conversion has been practiced for long; however, it is not used for
production of biofuels but used for direct heat and power applications.  
b. Gasification is a rapidly developing technique that is being developed for biofuels.

Gasification is based on the partial oxidation of the biomass in order to convert it into
a mixture of gases predominantly containing carbon monoxide and hydrogen known
as syngas. Syngas can be further processed downstream, after cleaning and
conditioning.
The biochemical route for biofuels production involves fermentation that converts biomass
sugars, such as glucose, fructose and sucrose, into alcohols primarily ethanol and carbon
dioxide (CO2). It comprises of biomass pre-treatment followed by fermentation of the sugars
to ethanol and finally separation and purification to yield pure etanol. (Ullah et al., 2015)
2.1 Agro residues

The residues were selected according to the generation volume (ton/year) in Mexico i.e.,
maize, barley, sorghum and wheat.
2.2 Data recollection
The data of the generation of agriculture residues were obtained from the Mexican official
records in INEGI (National Institute of Statitistics, Geography and Informtics, Mexico) from
2005 to 2014. The information was organized per year and state through the ten years. The
database was validated by using information per municipality and state agriculture
department.
2.3 Energy potential estimation
All results were analysed through the ten years, for final values the average value was
calculated.
The theoretical potential (TheP)
The theoretical potential is defined as the global energy that could be produced using the total
amount of agriculture residues.
Equation 1 is used to calculate the theoretical potential of the harvest as well as the process
residues, for each residue type a different residue-to-product ratio was used.
TheP =Y*RPR *A*(1−MC)*HHVdry (1)
148
Where: 
TheP: Theoretical potential of the harvest residues in GJ
Y: Yield of main produce in tfm/ha. 
RPR: Residue-to-product ratio of harvest residues 
A: Production area in ha 
MC: Moisture content as a percentage of the fresh matter in percentage
HHVdry: Higher heating value of the residues in GJ/tdm.
Sustainable potential
The sustainable potential is calculated using Equation 2 for the harvest residues, the
conservation of organic matter as residues is considered in order to maintain the carbon
content in soil and for avoiding the soil erosion. The Sustainable potential was calculated per
state and globally.
If S > E: SuP =TheP −(S*A*HHVdry ) (2)
If S < E: SuP =TheP −(E*A*HHVdry )
Where:
SuP: Sustainable potential of the harvest residues in GJ
TheP: Theoretical potential of the harvest residues in GJ
S: Resiudes required to maintain 2% SOC in t/ha
E: Soil cover required to reduce erosion to 10% of the bare soil erosion in t/ha
A: Production área in ha
HHVdry: Higher heating value of the residues in GJ/t
Technical potential
This potential is the residual value obtained considering the probable use that the residues
could have by using for feeding the livestock.
The technical potential was calculated using Equation 3 for the harvest of the residues
TechP = SuPt − (DCU* HHVdry ) (3)
Where: 
TechP: Technical potential of the harvest residues in GJ 
SuPt: Sustainable potential of the harvest residues in GJ. 
DCU: Demand by competing uses in t 
HHVdry: Higher heating value of the residues in GJ/t.
The competitive use of residues was estimated considering the size of the livestock population
per year and per state, and the demand for feed and bedding during the winter period whent
the animals cannot graze in the field (SAGARPA, 2014)
Maize residues (MzR)
The equation 4 was used to calculate the yearly demand of maize residues for livestock
feeding.
ARMR = (365-G)*0.25*Pp*Rd*Cc (4)
149
Where:
ARMR: Annual requirement of maize residues in kg
G: Days of grazing annually
Pp: Percentage of livestock fed with residues
Rd: Daily requirement of maize residues in kg
Cc: Cattle population
* The Rd value used was 12.5 kg/day according to the data reporteb by Gallagher, 2003.
Wheat residues (WhR)
The amount of wheat residues was estimated using the equation 5
RART = (365-p)*[(Cc*Pw*1.5)+(Co*Pw*1)+(Cp*Pw*0.5)]/1000 (5)
Where:
ARTR: Annual requirement of maize residues in kg
P: Days of grazing annually
Cc: Cattle population
Pw: Percentage of the livestock fed with wheat residues
Co: Ovine population
Cp: Porcine population

3.1. Generation of agriculture residues
The national average of cultivated area and residues generation were calculated (Figures 1
and 2); the yield per cultivar was calculated as well, for the lapse on study. The final value is
the result of the addition of all the cultivated areas and the yield per cultivar per each Mexican
state.
3%
18% Maize
Wheat
Sorghum
Barley
7%
72%
Figure 1. Average of the cultivated area per cultivar in Mexico (2005-2014)
150
4%
Maize
15% Sorghum
Wheat
36% Barley
45%
Figure 2. Average of the generation of residues in Mexico (2005-2014)
It is possible observe the distribution (cultivated area and yield) for the most important plants
cultivates in Mexico, however, some other cultivars were not considered, which are relevant
for some regions i.e., sugarcane in the south of Mexico. It does not exist relationship between
the cultivated area and the yield of cultivars.
3.2. Estimation of the theoretical, sustainable and technical potentials for the
agriculture residues
Global theoretical potential
The average of the theoretical and sustainable potential (PJ/year) are shown un the figure 2,
as has been reported, the sustainable value is lower than the theoretical potential. The
potentials were calculated for the states which cultivate the four plants i.e. maize, sorghum,
barley and wheat.
151
TheP
SuP
85000000
65000000
Energy potential (PJ)
45000000
25000000
5000000
Chihuahua
Tamaulipas
Hidalgo
Coahuila
Puebla
Baja California
Durango
México
Michoacán
Querétaro
Guanajuato
Jalisco
Sonora
San Luis Potosí
Nuevo León
Veracruz
Zacatecas
Oaxaca
-15000000
State
Figure 3. Values for the theoretical potential for the states which cultivate the four plants.
TheP and SuP values
The theoretical and sustainable potentials were estimated per year, the sustainable potential
was obtained considering the conservation of residues for maintaining the C concentration in
soil and avoiding the soil erosion, thus the sustainable values were lower than the theoretical
potentials. The consideration of the conservation of residues is crucial for maintaining the
productivity of the agricultural soil. Benefits of the residue conservation in agricultural soil
erosion and fertility decline have arisen throughout this agro-ecological zone (Scherr and
Yadav, 1996). The major limitations of the region are periodic drought, periodic excess
moisture and wind and water erosion. In the Mexican highlands, currently there is a
predominance of continuous maize monoculture wherever moisture permits, and mechanical
tillage is widespread, at least for land preparation prior to seeding. Although modest fertilizer
input appears to be used, cereal grain yield is low (<3 t/ha), fields are often weedy and nitrogen
deficient, soil structure is poor, and sheet and gully erosion is widespread (Fischer and Vidal.,
2002). The Figure 4 shows values for the theorethical potential between 700 and 900 PJ/year
and from 100 PJ/year to 300 PJ/year for the sustainable potential.
Figure 4. Theoretical an sustainable potentials (2005-2014)
152
Technical potential for maize and wheat residues
The technical potential was calculated considering four different scenarios and it was
estimated assuming the use of the maize (figure 5) and wheat residues (figure 6) for feeding
the livestock, it was calculated per year. The percentages of livestock considered were 25%,
50%, 75% and 100%. Not only in Mexico but in some asian, african and latinamerican
countries, the mixed crop–livestock farming systems prevail– typically rain-fed and
smallholder systems based on maize and ruminants and spanning diverse agro-ecologies.
Maize crop residues (stover) are an important by-product, primarily for feed use, often through
in situ stubble grazing and/or as ex situ forage (Hellin et al., 2013); same use has the wheat
residues and some others (Bogale and Yami, 2008). The higher values for the technical
potential of the wheat residues ((Figure 5) compared to the maize residues (Figure 6) is
explained by the incresing of the wheat yields in Mexico, which represent an important
measure of breeding; the wheat yield has increased by 25% over the past two decades (Lobell
et al., 2005). The Figures 5 and 6 show the possible availability scenarios for maize and wheat
residues, for both maize and wheat residues the values were not possitives after 25% of use
for feeding the livestock, however for some northern states i.e., Baja California and Sonora.
5000
0
2004 2005 2006 2007 2008 2009 2010 2011 2012 2013 2014
-5000
TechP
TechP - 25%
TechP - 50%
-10000
TechP - 75%
TechP - 100%
-15000
-20000
-25000
-30000
Año
Figure 5. Technical potential values for maize residues calculated considering different percentages
of livestock feeding.
153
50
0
2004 2005 2006 2007 2008 2009 2010 2011 2012 2013 2014
-50
-100
-150
TechP
TechP - 25%
TechP - 50%
TechP - 75%
TechP - 100%
-200 Año
Figure 6. Technical potential values for wheat residues calculated considering different percentages
of livestock feeding.
The HHV for the residue, the humidity conditions, and the RPR, are key factors which
determine the energy duty of the global volume. These factors are much more important than
the yield of the residue and its cultivated area. The maize, wheat and sorghum are the most
important cultivars in mexican agricultural soil, however, there are some other residues such
as sugarcane bagasse, cotton, coffee and rice, which in other mexican regions are
predominants.
4. CONCLUSIONS
 It was estimated the theoretical potential for agriculture residues generated in Mexico
(2005-2014), the sustainable and technical potential were calculated also. The use of
biomass as a source of energy in Mexico represents an sustainable alternative for
generating energy and reducing the impact of the use of fossil fuels.
 It was possible to determine the geographical distribution of these residues, its annual
generation, and the regional potentials.
 It will be necessary analyze the best methodology and technology for the use of the
biomass; the impact of its use on the environment should be studied.
ACKNOWLEDGMENTS
We thank to the Dirección de Apoyo a la Investigación y Posgrado (DAIP) of the Universidad
de Guanajuato, Mexico, for the support that we received for assisting to the First International
Conference on Bioenergy and Climate Change. EVN would like to thank to BV for her technical
assistance.
REFERENCES
Alemán-Nava, G. S., Casiano-Flores, V. H., Cárdenas-Chávez, D. L., Díaz-Chavez, R.,

Scarlat, N., Mahlknecht, J., ... & Parra, R. (2014). Renewable energy research
progress in Mexico: A review. Renewable and Sustainable Energy Reviews, 32, 140-
153.
154
Bogale, S., Melaku, S., & Yami, A. (2008). Potential use of crop residues as livestock feed
resources under smallholder farmers conditions in Bale highlands of Ethiopia. Tropical
and Subtropical Agroecosystems, 8(1), 107-114.
Cherubini, F. (2010). The biorefinery concept: using biomass instead of oil for producing
energy and chemicals. Energy Conversion and Management, 51(7), 1412-1421.
Fischer, R. A., Santiveri, F., & Vidal, I. R. (2002). Crop rotation, tillage and crop residue
management for wheat and maize in the sub-humid tropical highlands: I. Wheat and
legume performance. Field Crops Research, 79(2), 107-122.
Ghilardi A, Riegelhaupt E, Saldañaa R. Bioenergy resources in Mexico (Los Recursos
Bioenergéticos en México). In: Masera O, editor. Bioenergy in Mexico (La Bioenergía
en México). Mexico City: Mundiprensa; 2005 [in Spanish].
Gurung, A., & Oh, S. E. (2013). Conversion of traditional biomass into modern bioenergy
systems: A review in context to improve the energy situation in Nepal. Renewable
energy, 50, 206-213.
Hellin, J., Erenstein, O., Beuchelt, T., Camacho, C., & Flores, D. (2013). Maize stover use and
sustainable crop production in mixed crop–livestock systems in Mexico. Field Crops
Research, 153, 12-21.
Johnson, J. M., Coleman, M. D., Gesch, R., Jaradat, A., Mitchell, R., Reicosky, D., & Wilhelm,
W. W. (2007). Biomass-bioenergy crops in the United States: A changing paradigm.
Lobell, D. B., Ortiz-Monasterio, J. I., Asner, G. P., Matson, P. A., Naylor, R. L., & Falcon, W.
P. (2005). Analysis of wheat yield and climatic trends in Mexico. Field crops research,
94(2), 250-256.
Ramıŕ ez, A. M., Sebastian, P. J., Gamboa, S. A., Rivera, M. A., Cuevas, O., & Campos, J.
(2000). A documented analysis of renewable energy related research and
development in Mexico. International journal of hydrogen energy, 25(3), 267-271.
Scarlat, N., & Dallemand, J. F. (2011). Recent developments of biofuels/bioenergy
sustainability certification: A global overview. Energy Policy, 39(3), 1630-1646.
Scherr, S. J., & Yadav, S. (1996). Land degradation in the developing world: implications for
food, agriculture, and the environment to 2020.
Ullah, K., Sharma, V. K., Dhingra, S., Braccio, G., Ahmad, M., & Sofia, S. (2015). Assessing
the lignocellulosic biomass resources potential in developing countries: A critical
review. Renewable and Sustainable Energy Reviews, 51, 682-698.
Valdez-Vazquez, I., Acevedo-Benítez, J. A., & Hernández-Santiago, C. (2010). Distribution
and potential of bioenergy resources from agricultural activities in Mexico. Renewable
and Sustainable Energy Reviews, 14(7), 2147-2153.
World Bank Group (Ed.). (2014). World Development Indicators 2014. World Bank
Publications.
155
The activities on thermal degration of biomass at the

Biomass Researche Centre, Perugia, Italy
Pietro Bartocci*, Gianni Bidini, Paolo Laranci, Michele D'Amico, Francesco Fantozzi
Department of Engineering, University of Perugia, Perugia, Italy
*Corresponding author: bartocci@crbnet.it, Tel. +39 0755853773, Fax: +39 0755153321
ABSTRACT
Combustion is the traditional way to produce energy from residual and dedicated biomasses.
Together with combustion also gasification and pyrolysis are thermal degradation techniques,
which are used to produce heat and electricity in cogeneration plants. In this paper the main
facilities of the Biomass Research Centre (University of Perugia, Italy) are presented.
Combustion is analyzed and tested in a test bench for pellet boilers and stoves based on EN
303-5:2012 and EN 14785:2006. Gasification is performed in a downdraft reactor which is
coupled with a 6 kWe engine. While pyrolysis is studied both at laboratory scale and at pilot
scale with a 70kWe CHP plant, based on a rotary kiln reactor. Biomasses are firstly
characterized with proximate and ultimate analysis and then calorimetry is performed.
Evaluating mass and energy balances, energy plant efficiency can also be calculated.
Keywords: Combustion; Gasification; Pyrolysis
1. INTRODUCTION
Residential heating accounts for approximately 25-30% of the total energy demand in Italy.
The use of renewable energies in this sector may represent an important opportunity to
contribute to environmental sustainability and to a diversified energy system. The energy
market for buildings heating sees the lignocellulosic biomass in a position of great
competitiveness with fossil fuels. In the domestic sector, stoves, fireplaces and boilers of the
power of tens or hundreds of kW are currently marketed. However, at present, the Italian "fleet"
is represented largely by outdated and inefficient stoves. Recently the supply of
environmentally friendly combustion technologies have reached levels of efficiency, reliability
and comfort very similar to those of traditional systems for gas or oil, with about 90% efficiency,
compared to the old fireplaces and stoves. Several studies and surveys at national and
European level (Fiedler 2004, van Dam et al. 2009) show that the consumption of biomass for
domestic heating is increasing, especially the consumption of pellets, and the number of
installed units has almost doubled in the last 10 years. Being the actual market in continuous
expansion, there is a need for certification of biofuels and heating equipment. Dealing in
particular with heating equipment, at the moment, the market is pushing through small capacity
and low cost solutions, both for boilers and stoves that are often assembled and realized by
small enterprises and sold through the great distribution channels. For this reason it becomes
important to grant high performances, emissions and safety levels to align them to the existing
national and international standards. Biomasses and wastes are renewable energy sources
and their energy conversion does not increase the concentration of greenhouse gases in the
atmosphere. State of the art technology for electricity production from biomass is direct
combustion and thermal recovery in steam power plants, on a scale higher than 1 MWe, which
implies high investment and operational costs (due to the complexity of the supply chain). The
production of electricity on a micro-scale, on the other hand, presents an easier authorization
156
path and a simpler supply chain management, also increasing the possibility of using waste
heat for CHP applications. Gas turbines and ICEs are available on the market also on the
microscale but require the conversion of biomass into a liquid or gaseous which, may be
realized through pyrolysis and gasification processes (Moriconi et al. 20015).
Fixed bed downdraft technology is the most adopted and produces a gas having a low tar
content and the investment cost of the plant is about 3,000 €/kWe (Moriconi et al. 2015).
However this technology is not completely mature, with few companies selling micro-CHP
plants based on gasification process, with interesting payback of the investment and adequate
installed capacity and running hours, to minimize risk assessment. Syngas use in engines and
gas turbines, in fact, requires handling a gas with low LHV, tar, water and particulate
contamination; moreover different fuel gases in the mixture will show different combustion
behaviour (flame speed, ignition delay, etc.). This turns into costly gas cleaning systems and
combustion chamber modifications. From these premises a low cost a micro-scale gasification
plant was designed and built at the Biomass Research Centre of the University of Perugia to
be coupled to a small ICE or a botton cell SOFC for experimental testing. Given this
background, the authors have also proposed, and optimized, an innovative integrated
approach, namely the IPRP (Integrated Pyrolysis Regenerated Plant, see D'Alessandro et al.
2013), that combines a rotary kiln pyrolyzer and a gas turbine fuelled by the pyrolysis gas
produced from the thermal degradation of biomass and waste. This papers resumes the
results obtained and the path followed in the three above mentioned thermal degradation
processes facilities.

Combustion tests are performed in a test bench of the Biomass Research Centre (see figure
1). This was described in detail in Fantozzi et al. 2010.
Figure 1. Pellet boiler test bench layout
A pellet boiler produced by CLAM, model Vittoria of the thermal power of 20 kW, or a pellet
stove produced by Cola Ferroli can be used in experimental tests. Pyrolysis tests are
performed in a batch pyrolysis plant of the Biomass Research Centre, the layout is shown in
157
figure 4, also reported in Bidini et al. 2015. Test in the batch plant can be scaled up to the
dimensions of the IPRP technology.
Figure 2. Pyrolysis batch reactor
Figure 3. The IPRP concept layout
The IPRP technology is mainly composed of a Gas Turbine (GT) fuelled by synthesis gas,
obtained by slow pyrolysis of biomass and/or wastes in an externally heated reactor. The
energy required to sustain the pyrolysis process is provided by the exhaust gases of the GT
and by combustion of pyrolysis volatiles (tars) and solid (char) byproducts. Figure 3 shows the
flow diagram of the IPRP concept which considers two possible heat recovery schemes to
increase plant performance by means of regenerating the heat of the Joule cycle through a
double stage air preheating. Thermal energy of exhaust gases at the turbine outlet may be
recovered in the regenerator (REG) while thermal energy of the exhaust gases at the pyrolyzer
outlet may be recovered in the recuperator (REC). A preliminary thermodynamic optimization
of the IPRP technology was carried out through software modeling and a sensitivity analysis
on main design parameters such as GT compressor ratio (b), Regeneration Ratio (RR),
Turbine Inlet Temperature (TIT) and pyrolysis Temperature (Tp) which have a competing
effect on overall efficiency.
158
Gasification tests are performed at the Biomass Research Centre in a downdraft reactor with
capacity of 20 kg/h of chipped wood. This was presented in Moriconi et al 2015. The layout is
reported in figure 4.
Figure 4. Gasification plant layout (scrivere scrubber e poi!!! filtri e differenziarli)

The characteristics of the stove described and analyzed in the combustion test bench at the
biomass Research Centre Perugia (Italy) are proposed in table 1.
Table 7. Characteristics of the stove object of the study (data from the technical specification of the
stove)
Characteristics Description
CE Mark Present
Norm EN 14785 : 2006
Machine definition Residential space heating appliances fired by wood
pellets
Structure Steel and cast iron
Combustion chamber Steel and cast iron
Nominal thermal power 6.73 kW
Hourly consumption 1.56 kg/h
Efficiency 90.98 %
Flue gases temperature 155.9°C
Minimal safety distance 200 mm back wall
300 mm right/left wall
0 mm base
Dimensions H x L x P = 981 x 430 x 505 mm
Weight 93.5 kg
Capacity 25 liters (about 18 kg)
The pellet stove and pellet boiler test bench has been tested with pellet of different qualities.
Starting from different raw materials, such as: dedicated crops (Arundo Donax), waste
biomass (drifted wood) and pretreated biomass (biochar) pellet was produced and tested in
the test bench (see figure 5).
In all the cases biomass was firstly dried and then milled to the dimensions of 1 cm. Electricity
consumption was measured both during milling and pelletizing (done with a laboratory
machine). Once it was produced, the pellet was characterized through proximate analysis
159
(performed with a thermobalance), ultimate analysis (performed with CHN analyzer) and
calorimetry. Also pellet durability has been measured. After characterization of the fuel, tests
in the pellet boiler can be performed, calculating the efficiency (through the measurement of
exhaust gases composition and their temperature and also sampling the temperature of inlet
and outlet water) and measuring dust emissions.
Figure 5. Pellet production, characterization and tests in the combustion test bench
The mass balance obtained during preliminary tests with the gasification system is proposed
in table 2.
Table 2. Gasification Mass balances In Experimental test (15% biomass moisture), taken from
Moriconi et al. 2015
Mass Value
INPUT
Biomass (kg/h) 5.00
Air (kg/h) 7.74
OUTPUT
Solid residue (kg/h) 0.50
Tar and water (kg/h) 1.47
Producer gas (kg/h) 10.77
Mass balance of the Integrated Pyrolysis Plant (IPRP) is shown in table 3.
Table 3. Pyrolysis mass balance, taken from D'Alessandro et al. 2013

Mass balance %
Biomass 100
Char 30
Bioloil and water 38
Pyrogas 32
160
4. CONCLUSIONS
The Biomass Research Centre of the University of Perugia has developed different facilities
to test thermal degradation processes. A combustion test bench for pellet stoves and boilers
has been developed according to EN 303-5:2012 and EN 14785:2006. This has been used to
analyze different pellet fuels and different thermal appliances. Dealing with power and heat
production from biomass, through CHP plants, two prototypal plants have been developed:
the Integrated Pyrolysis regenerated Plant (with electrical power of about 70 kWe) and a
downdraft gasification plant (with electrical power of 6 kWe). Both can produce also biochar
to be used as soil amendment.
REFERENCES
Bidini G, Fantozzi F, Bartocci P, D‘Alessandro B, D‘Amico M, Laranci P, Scozza E, Zagaroli
M, 2015, Recovery of precious metals from scrap printed circuit boards through
pyrolysis, Journal of Analytical and Applied Pyrolysis, 111:140–147
D’Alessandro B, D’Amico M, Desideri U, Fantozzi F, 2013, The IPRP (Integrated Pyrolysis
Regenerated Plant) technology: From concept to demonstration, Applied Energy 101:
423–431
Fantozzi F, Vescarelli M, Bartocci P, Buratti C, 2010, Certification of pellet stoves and boilers
based on en 14785 standard – description of the test bench and preliminary results,
18th European Biomass Conference and Exhibition, 3-7 May 2010, Lyon, France
Fiedler F, 2004, The state of the art of small-scale pellet based heating systems and relevant
regulations in Sweden, Austria and Germany, Renewable and Sustainable Energy
Reviews 8:201–221;
Moriconi N, Laranci P, D’Amico M, Bartocci P, D'Alessandro B, Cinti G, Baldinelli A, Discepoli
G, Bidini G, Desideri U, Cotana F, Fantozzi F, 2015, Design and preliminary operation
of a gasification plant for micro-CHP with internal combustion engine and SOFC,
Energy Procedia 81: 298 – 308
UNI EN 303-5:2012, CALDAIE PER RISCALDAMENTO - PARTE 5: CALDAIE PER
COMBUSTIBILI SOLIDI, CON ALIMENTAZIONE MANUALE O AUTOMATICA, CON
UNA POTENZA TERMICA NOMINALE FINO A 500 KW - TERMINOLOGIA,
REQUISITI, PROVE E MARCATURA
UNI EN 14785:2006, APPARECCHI PER IL RISCALDAMENTO DOMESTICO ALIMENTATI
CON PELLET DI LEGNO - REQUISITI E METODI DI PROVA
van Dam, J., Faaij, A.P.C., Lewandowski, I., van Zeebroeck, B., 2009, Options of biofuel
trade from central and eastern to western european countries, Biomass an
161
Biomass blending as a way to reduce nox emissions

during the combustion of biomass residues
Brice Piednoir1, Jean-Michel Commandré2*, Anthony Benoist3, Sarah El Fassi4, Gilles
Vaitilingom5
CIRAD, UPR BioWooEB, F-34398 Montpellier, France

1brice.piednoir@cirad.fr
2jean-michel.commandre@cirad.fr (*corresponding author)
3anthony.benoist@cirad.fr
4sarah.el-fassi@etu.utc.fr
5gilles.vaitilingom@cirad.fr
ABSTRACT
The use of biomass as a source of energy through combustion has been subject to growing
interest, notably to tackle the issue of climate change. This has caused an increasing demand
on woody biomass resources, and led to consider other biomass resources, such as
agricultural residues. Among the atmospheric pollutants emitted during biomass combustion
are nitrogen oxides (NOx: NO and NO2), mainly originating from biomass N-content in boiler
operating conditions. These gaseous compounds notably impact ambient air quality and are
involved in the acid rain phenomenon. If NOx emissions are quite low for wood combustion,
they can act as a disincentive to the use of agricultural residues, in which nitrogen content is
higher than for wood.The aim of this work was to study the effect of biomass blending on the
NOx emissions of biomass residues. Grape marc, olive pomace, pine bark, rice husk and
wheat and rape straw were used. Pellets of these materials, alone or mixed together, were
made and combustion experiments were carried out in a laboratory furnace. O2, CO2, CO and
NOx concentrations were monitored at the furnace outlet. A linear correlation was found
between NOx emissions and the N-content of biomass residues when burned alone. The NOx
emissions of blended pellets, however, could not be linearly correlated to their N-content only.
The influence of other biomass constituents than nitrogen on NOx emissions, such as
inorganic materials, was thus highlighted and analyzed thanks to several blends and data
analyses.
Keywords: biomass residues; combustion; NOx emissions; blend.
1. INTRODUCTION
The growing concerns regarding greenhouse gas emissions tend to increase the interest for
the production of energy from biomass. Wood has been and still is the main biomass resource
used in combustion, for its good combustion properties. But the growing pressure on this
resource in the energy domain, and for other uses as in the paper and construction industries
lead to consider other biomass resources for energy production, such as agricultural residues.
Among the compounds constituting lignocellulosic biomass, nitrogen can be considered a

minor element, with N-content usually lower than 2 % w/w. However, the nitrogenous
compounds (NOX and N2O) emitted during biomass combustion are a major challenge to face
when considering thermal valorization of biomass. NO, on which this study focuses, can
indeed cause the formation of ozone in the atmosphere and are also implied in the
phenomenon of acid rain and the formation of particulate matter.
162
N-content of agricultural residues (straws, husks, etc…) generally being higher than N-content
of woody biomass, increased NOX emissions can be observed from the combustion of these
types of residues compared to wood [1,2]. As a consequence, NOX emissions, because of
their impact on the environment and health, can constitute a serious impediment to the thermal
valorization of such biomass residues, and solutions have thus to be implemented to mitigate
their NOX emissions.
Technological solutions already exist to do so, such as air staging [3–7] or selective catalytic
reduction (SCR) and selective non-catalytic reduction (SNCR) [8–10]. In SNCR for example,
urea or ammonia is injected in the flue gas in order to reduce NO to N2 thanks to NH3 or its
related NHi radicals (Thermal De-NOX mechanism) [8,9]. This work proposes another
approach and focuses on the properties of the fuels used. Biomass blending could indeed be
a way to address the issue of NOX emissions, if synergies in the NOX formation and reduction
mechanisms can be identified based on the chemical composition of the fuels. In particular,
inorganic materials contained in biomass could play a role in such N-related mechanisms,
catalyzing the NOX formation or reduction in N2. In the literature, information is scarce on the
role of inorganic materials on NOX emissions, but the influence of alkali metals, especially
calcium, has been studied before. Depending on the authors, calcium was suspected to
increase [11,12] or decrease [13,14] NOX emissions. These diverging conclusions might be
attributed to differing experimental conditions. Some authors injected Ca-additives in the flue
gas [13] when other study the reactivity of chars towards NO [14] or use demineralized coals
[11].
This work focuses on identifying which compounds contained in biomass could influence NO
emissions, in order to implement synergies in NO reduction through appropriate blending of
biomass residues. Several biomass residues were selected according to their contrasting
composition in N, Ca, K, Si, and Cl, as these compounds are usually linked to the problems
occurring during biomass combustion. Pellets of these residues were made and burnt at lab
scale. O2, CO2, CO and NO concentration were monitored during the experiments. Relevant
blends of these residues were then made and tested in the same combustion conditions.
2.1. Types of biomass used

The biomass residues used in this study were selected based on their chemical composition.
Nitrogen being a key element regarding NOX emissions, two biomass residues known to have
relatively high fuel-N content were chosen: olive pomace (OP) and grape marc (GM), with
nitrogen content of 1.4 and 1.9 % (w/w, db) respectively. On the opposite, the four other
biomass residues selected have lower fuel-N content (see Table 8): wheat straw (WS), rape
straw (RS), rice husk (RH) and pine bark (PB). These last residues were selected due to their
remarkable contents in other mineral elements (Ca, K, Si, Cl, S) identified in the literature as
playing a role in the technical and environmental issues linked to biomass combustion. More
specifically, the role of calcium and potassium on the formation or reduction of NOx has been
subject to discussion before [11–14].
163
Table 8. Chemical composition of the biomass samples used

Pine Wheat Rape Rice Olive Grape Pine
Bark Straw Straw Husk pomace Marc Pellet
(PB) (WS) (RS) (RH) (OP) (GM)
Proximate
analysis
Humidity (%) 10.3 8.8 10.1 8.8 7.4 9.6 10
Ash 1.3 6.5 6.3 15.4 4.4 8.4 0.5

% db
VM 72.2 76.8 75.5 65.5 74.5 65.9 79.8
FC 16.3 7.9 8.1 10.4 13.7 16.0 9.7
Ultimate
C 53.71 46.4 46.8 42.4 50.1 54.5 50.7
H 7.2 7.0 7.3 5.8 7.0 7.3 6.9

% db
N 0.2 0.4 0.5 0.5 1.4 1.9 0.2
O* 38.1 43.0 42.5 43.9 37.7 33.0 42.3
Cl 276 640 2370 1386 2905 174
Si 921 21433 1296 63306 2591 2734
Al 1526 313 74 93 507 275
Fe 238 242 66 105 484 465
Ti 20 40 10 4 29 24
mg/kg (db)
Ca 3509 3402 10741 1647 6265 7534
Mg 293 609 541 604 2235 742
Na 50 143 598 497 197 119
K 657 4422 10947 6280 20137 16814
S 115 522 2504 418 1176 1328
P 137 355 799 571 1939 2345
Mn 97 63 11 198 19 29
A commercial pine pellet was also used as a reference for the implementation of the
experiment.
2.2. Samples preparation

Pellets of pure and blended biomass were prepared for this work. The pellet form was used:
1) to improve the contact between two different biomass types, promoting interactions during
the combustion; and 2) because it is a promising way to precondition biomass, tackling the
issue of its low bulk density.
164
Pellets were individually produced at lab scale, using an electromechanical press apparatus
equipped with a 10kN measure cell, and with the help of a cylindrical steel mold preheated at
100 °C.
The biomass residues selected were first ground thanks to a cutting mill through a 2 mm sieve.
About 500 mg of ground biomass or mix of ground biomass residues were then introduced in
the steel mold (6 mm diameter) and compressed to 8kN (~280 MPa), at a speed of 0.07 mm.s-
1
from 0 to 2kN and 0.05 mm.s-1 from 2kN to 8kN. The pressure was finally held for 20
seconds. The steel mold was then raised and the obtained pellet was pressed out of the
channel. These experimental conditions were chosen based on the literature [15–17] and
following several tests in order to obtain cohesive pellets. The resulting pellets were about 150
mm in height and 6mm in diameter. Thanks to this single pellet production, the composition of
the blended biomass pellets could be accurately controlled for each pellet to be burned.
2.3. Combustion apparatus

Combustion experiments were led in batch conditions at lab-scale using a setup represented
on Figure 23.
Figure 23. Schematic representation of the combustion setup.
The setup consisted of a stainless steel tubular reactor, horizontally inserted in an electrically
heated furnace made of refractory firebrick (4). The first part of the tubular reactor was cooled
down thanks to a cold water heat exchanger (3). An insertion probe equipped with a sample
holder (5) allowed the sample to be placed in the cold zone, before being inserted in the heated
zone. The sample holder was a crucible made of stainless steel mesh (200 µm). The
temperature in the reactor was measured and regulated thanks to a thermocouple placed at
mid distance in the heated zone, in the refractory firebrick wall close to the surface of the
stainless steel reactor. An additional heating resistance (6) was placed near the end of the
reactor, to maintain a quite homogeneous temperature distribution from the middle of the
reactor to its end.
At the entrance of the reactor, two mass flow controllers fed the process with N2 (1) and O2
(2). At the outlet of the reactor, the combustion gases were cooled down (7) in order to
165
condensate water and then cleaned through two filters: 1) a tubular filter containing silica gel
(8), to retain the remaining water traces in the gas, and filter out part of the unburnt particles
in the flue gas; and 2) a quartz filter (0.3 µm) (9) efficiently retaining the unburnt particles still
contained in the combustion gases. Once cooled down and cleaned, a representative part of
the gas was then sent to a gas analyzer (ABB EL 3020) for the quantification of O2, CO2, CO
and NO concentrations.
2.4. Procedure
For each experiment, one pellet of about 500 mg was placed in the sample holder in the cold
zone. The mass flow controllers delivered a gas flow rate of 3.8 Nl.min-1 with 21% O2 and 79%
N2. The reactor temperature was set up at 800 °C, and the additional heating resistance was
set up at 650°C (due to technical constraints). The pellet was inserted in the reactor in about
10 seconds, and placed 24 cm after the beginning of the heated zone. From this position to
the end of the reactor, temperature remains quite constant (~780°C). The combustion started
shortly after the introduction of the pellet and was monitored thanks to the gas analyzer that
pumped 1 l.min-1 of flue gas. Once the combustion was over (O2 concentration back to 21%;
CO2, CO and NO concentration null), the sample holder was brought back to the cold zone.
For each pellet (pure biomass or blend), at least three experiments were carried out.
The gas analyzer monitored the O2 and CO2 volume concentration in percentage, and the CO
and NO volume concentration in ppm. Following an experiment, the gas releasing curves
could be integrated according to Eq (1) (given as an example for NO):
𝑡 𝑄(𝑡)
𝑛𝑁𝑂 = ∫0 𝑓 𝐶𝑁𝑂 (𝑡) × 𝑉𝑚
𝑑𝑡 Eq. (1)
with 𝑛𝑁𝑂 in mol, tf the duration of the combustion in s, 𝐶𝑁𝑂 the volume concentration of NO in
the flue gas, Q(t) the flue gas flow rate in Nl.s-1 and Vm the molar volume of an ideal gas in
l.mol-1, here at 1 atm and 0 °C.
Q(t) is calculated thanks to the measured concentration of O2, CO2, CO and NO and the known
N2 flow rate:
𝑉̇𝑁2
𝑄(𝑡) = Eq. (2)
1−(𝐶𝑂2 +𝐶𝐶𝑂2 +𝐶𝐶𝑂 +𝐶𝑁𝑂 )
with 𝑉̇𝑁2 in Nl.s-1 and 𝐶𝑂2 , 𝐶𝐶𝑂2 , 𝐶𝐶𝑂 and 𝐶𝑁𝑂 the volume concentration of the gases.
The parameters introduced to quantify the NO emissions were:

𝑚𝑁𝑂
- the mass of NO released for 100 g of biomass (dry basis): 𝑚 (𝑑𝑏)
× 100
𝑏𝑖𝑜𝑚𝑎𝑠𝑠
𝑛𝑁𝑂
- the molar NO conversion rate (%): 𝑋𝑁−𝑁𝑂 = 𝑛𝑁𝑏𝑖𝑜𝑚𝑎𝑠𝑠 (𝑑𝑏)
× 100;
166
3.1. Characterization of the combustion experiments

On Figure 24 are presented the release profiles of CO2, CO and NO as well as the O2
consumption, for the combustion of a pellet of olive pomace.
a) b)
c)
Figure 24. O2, CO2 (a), CO (b) and NO (c) release profiles for the combustion of a pellet of olive
pomace.
The profiles presented on Figure 24 illustrate the typical results obtained for the batch
combustions carried out in this study. Two phases will be distinguished on these curves, as it
can be found in the literature [5,18,19]. The first phase, shorter and with fast reaction rates as
illustrated by the O2 consumption, is attributed to the dominance of the oxidation of compounds
released during the devolatilization stage. The second phase is attributed to the dominant
combustion of char. For NO emissions, the separation between the two phases was defined
at the beginning of the plateau of NO concentration in the flue gas (Figure 24.c)).
Low level of CO (<1000 ppm) were generally observed, even if pic emissions of ~5000 ppm
during the devolatilization stage were obtained for rice husk (RH), wheat straw (WS) and rape
straw (RS). The higher levels of CO emissions can be linked to the fact that, for our
experiment, the total combustion of these three residues is faster than for pine bark (PB), olive
pomace (OP) and grape marc (GM), respectively in about 200 seconds for the the RH, WS
and RS and in about 400 seconds for PB, OP and GM. The oxygen inlet being stable, a high
demand of oxygen can occur during the combustion of rice husk, wheat straw and rape straw,
causing local lacks of oxygen and thus increased emissions of CO due to an incomplete
oxidation of carbon to CO2. The integration of the CO2 and CO emissions curves however
reveals that the total emissions of CO represent less than 1.5% of the total emissions of CO2
and CO, whatever biomass residue sample is considered. This piece of information, coupled
167
to a carbon balance always superior to 90% (molar ratio of the amount of carbon emitted as
CO2 and CO to the initial amount of carbon in biomass), indicates that the combustions carried
out in our experiment occur in acceptable operating conditions and are almost complete. The
remaining carbonaceous materials lacking in the carbon balance are to be linked to soot
deposits in the reactor and in the gas path, though.
The temperature set for our experiments is 800 °C, but the combustion being exothermic,
higher temperatures can be reached in the reactor. Even if in the range 800 – 1200 °C, fuel-
N is supposed to be the main contributor to NOX emissions for biomass combustion compared
to thermal-NOX and prompt-NOX [20,21], such temperature levels may raise questions
concerning the influence of nitrogen contained in the inlet gas on NO emissions during the
experiments.
A test was thus carried out in an atmosphere composed of 79% Ar and 21% O 2 to determine
the amount of thermal-NOX produced during the combustion experiments. A combustion test
was performed with the reference pellet, the commercial pine wood pellet. Among our different
biomass samples, the commercial pine wood pellet was the one with the lowest fuel-N content
(0.19% d.b), and if thermal-NOX were to be formed, it would have been more noticeable for a
biomass which fuel-N content was low. No differences in NO emissions of commercial pine
wood pellet burnt in nitrogen or argon atmosphere could though be observed
(0.12 ± 0.01 gNO/100g biomass db in 79%N2/21%O2 atmosphere,
0.14 gNO/100g biomass db for the test in 79%Ar/21%O2 atmosphere), indicating that NO
emissions were strongly linked to fuel-N content in these operating conditions.
3.2. NO emissions of pure biomass pellets
Figure 25. NO emissions obtained for pure biomass pellets as a function of their N-
content
As can be expected, the NO emissions increase with an increased N-content of biomass. But
Figure 25 additionally reveals a strong linear correlation (R² = 0.989) between NO emissions
and fuel-N content for the six biomass residues samples used in this study:
gNO/100 g biomass db = 0.1866 × %N (d.b) + 0.0997 Eq (3)
168
The same linear trend can be observed for other results from the literature (Figure 26) on the
range 0-2 %N (d.b) [3,9,22] . Malt waste from Winter et al. [9], with a high N-content (4.99%
d.b), however digress from this linear correlation.
Figure 26. Comparison between the results obtained in this work and results from
the literature regarding NO emissions as a function of the N-content of biomass.
The NO emissions from literature were converted in gNO/100g biomass db. These works
indeed used the molar NO conversion rate XNO to express their results. Figure 27 shows the
conversion rate XNO compared to biomass-N content, for the combustion experiments led in
this work and in the literature.
Figure 27. Comparison between the results obtained in this work and results from
the literature regarding N to NO conversion as a function of the N-content of biomass.
169
Good agreement is also found between our results and the ones provided by the authors for
this criterion. It has already been observed that with an increased N-content in biomass, the
conversion of initial fuel N to NO decreased. It has to be noticed that despite the different
setups, including ours, the hyperbolic trend previously observed in the literature remains valid.
Pine bark (PB) is the biomass residue sample with the lowest nitrogen content (0.21% d.b)
and more than 32% of the initial N is emitted as NO, whereas for grape marc (GM), with a
nitrogen content of 1.89% d.b, only 11% of N is converted to NO. This phenomenon is
generally explained by the fact that with an increased N-content, more NHi radicals are
released during the devolatilization stage, promoting the reduction of NO to N2 through the
thermal-DeNOX mechanism.
3.3. NO emissions for the blend pine bark/grape marc

The strong linear correlation between the mass of NO emitted for 100 g of dried biomass and
the fuel-N content (Eq(3)) indicates that nitrogen content of biomass would largely explain NO
emissions for this criterion. In order to check the validity of this assumption, blended pellets of
pine bark (PB) and grape marc (GM) were prepared. These two biomass residues are indeed
the ones with an extremal N-content among our samples (0.2% and 1.9%N w/w db
respectively). Figure 28 shows the NO emissions obtained for the combustion experiments
led on these PB/GM blended pellets.
Figure 28. NO emissions obtained for the pure and blended pellets of pine bark
(PB) and grape marc (GM) regarding their N-content
As can be seen on Figure 28, the NO emissions of the blended pellets do not follow the
correlation found for pure biomass pellets. At least three experiments were carried out for each
blend. The gap between the NO emissions found for the blended pellets and the emissions
that could be expected if they had followed the linear correlation (Eq(3)) is higher than the
standard experimental deviation, supporting the conclusion that NO emissions of the blended
pellets significantly differs from the correlation found for pure biomass pellets. The most
significant gap is obtained for the blended pellet with a 0.9% (db) N-content, with
170
0.35 ± 0.01 gNO/100g biomass db emitted against 0.27 gNO/100g biomass db expected with
the linear correlation found in 3.2, i.e approximately 27% deviation. These deviations from the
expected NO emissions reveal an interaction between PB and GM during combustion. Figure
29 shows the average NO emission profile obtained for this blend, compared to the theoretical
profile calculated thanks to the weighted addition of the NO emissions obtained for pure pellets
of PB and GM.
Figure 29. Comparison between the experimental and theoretical NO release

profiles of the 0.9%N w/w db blended pellet of pine bark and grape marc
As shown on Figure 29, differences between the theoretical and actual NO emissions take
place both during the devolatilization phase and the char combustion phase. When integrating
the NO emissions, this difference is found to be twice more important during the
devolatilization phase. During this first combustion phase, 42% more gNO/100g biomass db
than expected are indeed emitted, whereas during the char combustion phase, 19% more
gNO/100g biomass db than expected are emitted. The difference during the char combustion
phase does not come from higher concentration of NO in the flue gas, but from a longer char
combustion. The same trend can be observed for the experiments carried out on the other
blended pellets of PB and GM. That indicates that gas-gas or solid-gas interactions increasing
the NO formation could take place during devolatilization stage for this blend.
Several hypothesis can be made to explain this behavior. The NH3/HCN ratio could be
decreased by adding GM to PB, modifying the composition of the volatile atmosphere during
the devolatilization stage, and thus decreasing the thermal De-NOX effect. Mineral matter
could also catalyze the formation of NO. Grape marc indeed contains important quantities of
potassium and phosphorous compared to pine bark Table 8, these two compounds being
known to easily volatilize. No information on such mechanisms could however be found in the
literature. On the opposite, the role of calcium in NO formation or reduction has been studied
before [11–14] but no consensus exists on these types of catalytic mechanisms, some authors
concluding that calcium could increase NOX emissions while others would find opposite
results. This mineral compound could play a part in the interactions between PB and GM, and
the resulting difference observed between the theoretical and actual NO emissions of the
blended pellets. The influence of calcium was thus analyzed thereafter.
171
3.4. Influence of calcium on NO emissions through biomass blending

In order to study the influence of calcium on NO emissions during biomass combustion,
blended pellets with a fix N-content and a variable Ca-content were first prepared. Rice husk
(RH) and rape straw (RS) were used. These two biomass residues have almost the same N-
content (~ 0.5% d.b), but RH contains about 5 times less Ca than RS. The NO emissions
obtained for the blended pellets of these two residues are shown on Figure 30.
Figure 30. NO emissions obtained for the pure and blended pellets of rape straw
(RS) and rice husk (RH) regarding their N-content
The experimental results for the blended pellets follow a linear trend between the NO
emissions of the pure biomass pellets. More specifically, the NO emissions obtained from the
combustion of the blended pellets do not differ from the NO emissions that could be expected
considering a weighted addition of the NO emissions from rape straw and rice husk. No
interactions between the two residues are thus observable. The difference in calcium
concentration between rice husk and rape straw may not be high enough to reveal any
deviation from the expected NO emissions through the blending of these residues. Zhao et al.
[11] for example compared the NOX emissions of a demineralized coal to the NOX emissions
of the same coal mixed with 2.5-7.5 % w/w of pure CaO or Ca(OH)2.
To further investigate the influence of calcium in possible interactions between biomass

𝐶𝑎
residues during their combustion regarding NO emissions, blended pellets with a fixed ratio
𝑁
were prepared. The biomass residues best answering to such a condition were rice husk and
𝐶𝑎
olive pomace, with 𝑁
ratios of 0.12 and 0.16 respectively. The NO emissions obtained for
these pellets are presented on Figure 31.
172
Figure 31. NO emissions obtained for the pure and blended pellets of rice husk (RH)
and olive pomace (OP) regarding their N-content
Similarly to the results obtained for the PB/GM blend in 3.3, the NO emissions of the blended
pellets of rice husk and olive pomace do not follow the linear correlation between NO
emissions and N-content (Eq(3)). This time, a reduction of NO emissions compared to the
expected emissions is observed Figure 31. The deviation is linked for this blend to decreased
NO emissions during the char combustion stage, as can be seen on Figure 32, on which is
plotted the NO emissions profile of the 0.8%N (d.b) blended pellet of RH-OP, compared to its
theoretical NO emissions profile. As shown on this figure (Figure 32), it seems that the blended
pellets fully burn in a shorter time than expected. The experimental plateau concentration of
NO in the flue gas during char combustion is however close to the theoretical value. As a
consequence, fewer nitrogen than expected is converted to NO during the char combustion
phase. This results in the decreased global NO emissions for the blended pellet observed on
Figure 31.
The experimental and theoretical NO emissions during the devolatilization phase overlap so
no interactions between rice husk and olive pomace can be observed for this type of blend in
𝐶𝑎
the devolatilization phase, as opposed to the PB/GM blend. The molar ratio 𝑁
being constant
𝐶𝑎
for the RH/OP blends whereas it varies in the PB/GM blends, 𝑁
could be an indicator of
possible interactions between biomass residues concerning NO emissions.
173
Figure 32. Comparison between the experimental and theoretical NO release

profiles of the 0.9%N w/w db blended pellet of rice husk and olive pomace.
4. CONCLUSIONS
NOX emissions of agricultural residues can be quite high compared to wood. In this study, NO
emissions of grape marc (0.46 gNO/100g biomass db) were for example almost 4 times higher
than the reference pine pellet (0.12 gNO/100g biomass db). Even if the conversion of N to NO
decreases with an increasing N-content in biomass, the NO emissions (in
gNO/100 biomass db) were found in this work to be linearly correlated to the N-content of
biomass in the range 0-2 %N (w/w db). The N-content of agricultural residues generally being
higher than the N-content of wood, this correlation confirms that greater NOX emissions have
to be expected when burning agricultural residues. Blended pellets of pine bark (0.2 %N, w/w
db) and grape marc (1.9 %N, w/w db) however revealed that other compounds than nitrogen
had to be taken into account when considering NOX emissions, as the linear correlation
between NO emissions and N-content was not valid for this blend. The influence of calcium,
controversial in the literature, was thus investigated through two other blends, but no clear
effect could be observed. Still, a NO reduction could be observed when blending rice husk
and olive pomace. The experiments carried out in this work revealed that synergies can exist
regarding NO emissions when burning blended pellets. Further research is still needed to
identify which compounds are at the origin of the phenomenon, but blending could in the future
be used to lower NO emissions of agricultural residues, reducing the need in other NO
reduction measures (air staging, SCR/SNCR etc…) and thus promoting the thermal
valorization of agricultural residues.
REFERENCES
[1] Werther J, Saenger M, Hartge E-U, Ogada T, Siagi Z. Combustion of agricultural
residues. Prog Energy Combust Sci 2000;26:1–27. doi:10.1016/S0360-
1285(99)00005-2.
[2] Khan A a., de Jong W, Jansens PJ, Spliethoff H. Biomass combustion in fluidized bed
boilers: Potential problems and remedies. Fuel Process Technol 2009;90:21–50.
doi:10.1016/j.fuproc.2008.07.012.
[3] Díaz-Ramírez M, Sebastián F, Royo J, Rezeau A. Influencing factors on NOX emission
level during grate conversion of three pelletized energy crops. Appl Energy
2014;115:360–73. doi:10.1016/j.apenergy.2013.11.011.
[4] Houshfar E, Skreiberg Ø, Todorović D, Skreiberg A, Løvås T, Jovović A, et al. NO x
emission reduction by staged combustion in grate combustion of biomass fuels and fuel
174
mixtures. Fuel 2012;98:29–40. doi:10.1016/j.fuel.2012.03.044.

[5] Giuntoli J, de Jong W, Verkooijen AHM, Piotrowska P, Zevenhoven M, Hupa M.
Combustion Characteristics of Biomass Residues and Biowastes: Fate of Fuel
Nitrogen. Energy & Fuels 2010;24:5309–19. doi:10.1021/ef100571n.
[6] Samuelsson JI. Conversion of Nitrogen in a Fixed Burning Biofuel Bed. Biomass 2006.
[7] Abelha P, Gulyurtlu I, Cabrita I. Release of nitrogen precursors from coal and biomass
residues in a bubbling fluidized bed. Energy and Fuels 2008;22:363–71.
doi:10.1021/ef700430t.
[8] Kasuya F, Glarborg P, Johnsson JE, Dam-johansen K. THE THERMAL DeNOx
PROCESS: INFLUENCE OF PARTIAL PRESSURES AND TEMPERATURE
1995;50:1455–66.
[9] Winter F, Wartha C, Hofbauer H. NO and N2O formation during the combustion of
wood, straw, malt waste and peat. Bioresour Technol 1999;70:39–49.
doi:10.1016/S0960-8524(99)00019-X.
[10] Skalska K, Miller JS, Ledakowicz S. Trends in NOx abatement: A review. Sci Total
Environ 2010;408:3976–89. doi:10.1016/j.scitotenv.2010.06.001.
[11] Zhao Z, Li W, Qiu J, Wang X, Li B. Influence of Na and Ca on the emission of NOx
during coal combustion. Fuel 2006;85:601–6. doi:10.1016/j.fuel.2005.09.001.
[12] Ren Q, Zhao C, Duan L, Chen X. NO formation during agricultural straw combustion.
Bioresour Technol 2011;102:7211–7. doi:10.1016/j.biortech.2011.04.090.
[13] Lissianski V V., Zamansky VM, Maly PM. Effect of metal-containing additives on nox
reduction in combustion and reburning. Combust Flame 2001;125:1118–27.
doi:10.1016/S0010-2180(01)00231-0.
[14] Zevenhoven R, Hupa M. The reactivity of chars from coal, peat and wood towards NO,
with and without CO. Fuel 1998;77:1169–76. doi:10.1016/S0016-2361(98)00023-4.
[15] Rhén C, Gref R, Sjöström M, Wästerlund I. Effects of raw material moisture content,
densification pressure and temperature on some properties of Norway spruce pellets.
Fuel Process Technol 2005;87:11–6. doi:10.1016/j.fuproc.2005.03.003.
[16] Kaliyan N, Vance Morey R. Factors affecting strength and durability of densified
biomass products. Biomass and Bioenergy 2009;33:337–59.
doi:10.1016/j.biombioe.2008.08.005.
[17] Stelte W, Sanadi AR, Shang L, Holm JK, Ahrenfeldt J, Henriksen UB. Recent
developments in biomass pelletization - a review. BioResources 2012;7:4451–90.
[18] Bai J, Yu C, Li L, Wu P, Luo Z, Ni M. Experimental Study on the NO and N 2 O Formation
Characteristics during Biomass Combustion. Energy & Fuels 2013;27:515–22.
doi:10.1021/ef301383g.
[19] Houshfar E, Wang L, Vähä-Savo N, Brink A, Løvås T. Characterisation of CO/NO/SO2
emission and ash-forming elements from the combustion and pyrolysis process. Clean
Technol Environ Policy 2014;16:1339–51. doi:10.1007/s10098-014-0762-3.
[20] Glarborg P, Jensen a. D, Johnsson JE. Fuel nitrogen conversion in solid fuel fired
systems. Prog Energy Combust Sci 2003;29:89–113. doi:10.1016/S0360-
1285(02)00031-X.
[21] Obernberger I, Brunner T, Barnthaler G. Chemical properties of solid biofuels—
significance and impact. Biomass and Bioenergy 2006;30:973–82.
doi:10.1016/j.biombioe.2006.06.011.
[22] Miranda T, Arranz JI, Montero I, Román S, Rojas CV, Nogales S. Characterization and
combustion of olive pomace and forest residue pellets. Fuel Process Technol
2012;103:91–6. doi:10.1016/j.fuproc.2011.10.016.
175
Copper bioleaching from waste printed circuit

boards by Acidithiobacillus ferrooxidans
Eva Benzal, Montserrat Solé, Concepció Lao-Luque, Xavier Gamisans, Antonio David
Dorado*
Department of Mining, Industrial and ICT Engineering, Universitat Politècnica de Catalunya, Av.
Bases de Manresa 61-73, 08242 Manresa, Spain
Corresponding author: toni.dorado@upc.edu , Tel. +34 938777353, Fax: +34 938777286.
ABSTRACT
The work presented in this manuscript is focused on the bioleaching of copper in printed circuit
boards (PCB) from mobile phones by Acidithiobacillus ferrooxidans. In particular, the effect of
solid-liquid ratio of PCB and mineral medium and the effect of Fe(II) and Fe(III) on the process
has been investigated. For the first time, the effect of e-waste concentration on the process
efficiency has been analysed distinguishing between the chemical and the biological metal
leaching. The influence of pH control in copper bioextraction was evaluated. Moreover, the
study presented herein compares the bioleaching process of a mineral sample with an
electronic waste sample. Results show that when there are microorganisms involved in the
process, up to an increase of 30% of copper is recovered from the PCBs extracting the 99%
of copper contained in the boards. However, the leaching rate depends on the amount of
waste treated. It has been proved that iron ions are responsible of the chemical process,
having similar effects on copper recovery when Fe(III) was used at the beginning of the
leaching process instead of Fe(II).. Finally, it has been demonstrated that it is more efficient
extracting copper from electronic waste than from the process using chalcopyrite.
1. INTRODUCTION
Electronic waste (e-waste) consisting of discarded computers, televisions, cell phones,

stereos and other electronic equipment, has become a major problem worldwide. Rapid
advances in technology mean that electronic products are becoming obsolete in a short time.
This, coupled with explosive sales in consumer electronics, means that more products are
being disposed, even if they still work. In 2014, more than 40 million tons of e-waste were
produced in the world (Baldé, Wang, Kuehr, & Huisman, 2015). One of the main problems
with this type of e-waste is that they are packed with toxic chemicals — arsenic, lead, mercury,
poly-brominated flame retardants among others. On the other hand, e-waste also contain an
important part of valuable metals such as copper, silver or gold, which could be retrieved to
be used again (Fornalczyk, Willner, Francuz, & Cebulski, 2013). In general, chemical methods
are employed to recover these metals from mineral matrix (ores) or plastic matrix (e-waste).
Lately, bioleaching has been tested as an alternative to chemical processes showing important
advantages such as low cost, high efficiency and environmental friendliness (Arshadi &
Mousavi, 2015; Chen, Yang, Liu, Dong, & Liu, 2015).
Bioleaching is based on the release of metals from metal concentrates, as e-waste, using the
microorganism’s activity. Regarding to the microorganisms, chemoautotrophic
microorganisms are one of the most useful to bioleach (Mražíková, Marcinčáková, Kaduková,
& Velgosová, 2013; Joanna Willner & Fornalczyk, 2013). They are aerobic and grow in very
176
acidic conditions of pH 1.5 – 3.0 (Choi, Cho, Kim, & Kim, 2004). In bioleaching, the
microorganism has an indirect role on copper extraction. In fact, copper is oxidized by Fe(III)
ions which in turn is reduced to Fe(II). The task of the microorganism is the oxidation of Fe(II)
ions to Fe(III), regenerating the agent responsible of copper bioleaching (Chen et al., 2015).
According to these authors, an iron cycle occurs during the leaching process of copper,
following Eq. (1) and (2).
𝐶𝑢 + 2 𝐹𝑒 3+ → 𝐶𝑢2+ + 2 𝐹𝑒 2+ (1)
𝐵𝑖𝑜𝑙𝑜𝑔𝑖𝑐𝑎𝑙 𝑎𝑐𝑡𝑖𝑣𝑖𝑡𝑦
4 𝐹𝑒 2+ + 4 𝐻 + + 𝑂2 → 4 𝐹𝑒 3+ + 2 𝐻2 𝑂 (2)
An important factor to take into account in bioleaching process is the amount of e-waste
treated. Some authors have focused on this parameter (Adhapure, Waghmare, Hamde, &
Deshmukh, 2013; Yang et al., 2014; Zhu et al., 2011). All of them concluded that best results
are obtained when the concentration of e-waste treated is lower than 15 g/L. However, none
of these studies have taken into account the possible chemical leaching, considering the whole
copper extraction due to the biological process.
Besides e-waste concentration, another important factor influencing the leaching process is
the medium pH. As stated previously, the microorganisms which are involved in bioleaching
process grow in very acidic conditions (pH 1.5 – 3.0). Therefore, to assure a proper biological
process it would be important to maintain acid pH over the whole experiment. Some authors
studied the pH influence on the bioleaching process (Bas, Deveci, & Yazici, 2013; Shah, Tipre,
Purohit, & Dave, 2015; Joanna Willner & Fornalczyk, 2013), despite they did not compare the
process performed under controlled and uncontrolled pH.
The aim of the work presented herein is to evaluate some operational aspects of the extraction
of copper from printed circuit boards by means of the biological activity of Acidithiobacillus
ferrooxidans, and to identify the key parameters in the process. The role of Fe(II) and Fe(III),
the effect of e-waste concentration on the process efficiency distinguishing between the
chemical and the biological leaching and the comparison between copper bioleaching from
mineral ore and from e-waste was tested.
2.1. Electronic scrap
The waste printed circuit boards (PCB) used in this work comes from end-of-life mobile
phones. For experimental use, the scraps were sheared after manually removing the main
electronic components such as resistors, capacitors and chips among others. Then, they were
crushed and sieved, collecting the particles between 0.2 and 1.0 mm of diameter.
In order to determine the copper content in the PCB 0.15 g samples of electronic waste was
digested with 10 mL of HNO3 : HCl (3:1) at 150 ºC for 15 minutes in a microwave apparatus
(Microwave System, Millestone, Italy). The digestate was diluted to a known volume and
analysed by atomic absorption spectrometry (Solar S2, Thermo Scientific, United States). The
whole procedure was repeated 5 times to assess their repeatability. The average content of
Cu was 34.5% and the coefficient of variation was lower than 2.6%.
177
2.2. Mineral sample
The mineral used in this work was a chalcopyrite ore from La Negra’s mine in Querétaro
(Mexico). It was crushed and sieved, collecting the particles between 0.2 and 1.0 mm of
diameter. To determine its copper content, the ore was analysed by atomic absorption
spectrometry (AAS) after acid digestion using a microwave apparatus (Microwave System,
Millestone, Italy). The procedure to analyse it was the same procedure used to analyse the
electronic scrap, explained above. In this case, the average content of Cu was 26.4% and the
coefficient of variation was 1.5%.
2.3. Microorganisms and mineral medium
In this study an Acidithiobacillus ferrooxidans (DSMZ 24419) strain was used. It was provided
by Leibniz-Institut DSMZ-Deutche Sammlung von Microorganismen and Zelkulturen from
Germany. The mineral medium used in the experiments has prepared as follows: (NH4)2SO4
2.00; K2HPO4 0.50; MgSO4 · 7 H2O 0.50; KCl 0.10; Ca(NO3)2 · 4 H2O 0.014; grams were
dissolved in 950 mL of distillate water. The pH was adjusted with H2SO4 10 N to pH 2. Then it
was autoclaved at 121 ºC and 1 atm during 15 minutes for sterilization. On the other hand,
8.00 grams of FeSO4 · 7 H2O were dissolved in 50 mL of distillate water and sterilized by
filtration through a 0.22 µm membrane filter. After that, both sterile solutions are mixed.
The study of ferric ion effect was carried out modifying the mineral medium describe
previously. In this case, the amount of ferrous ion concentration was replaced by ferric ion
concentration. In particular, the 8.00 g/L FeSO4 · 7 H2O was substituted by 5.75 g/L Fe2(SO4)3
· X H2O in order to maintain the same moles of ferric and ferrous ions. The rest of the salts
were in the same concentration and the sterilization was done following the same procedure.
2.4. Bioleaching experiments
Bioleaching experiments were performed in 500 mL Erlenmeyer flasks containing 180 mL of

mineral medium, 20 mL of inoculum and different concentrations of PCB, depending on the
effect studied. Abiotic control experiments were also carried out. The flasks were kept at 30ºC
and 120 rpm during 23 days, taking samples of the leaching solutions every 3-4 days to
analyse them.
Total iron and copper ions concentration in the bioleachates were analyzed by an atomic
absorption spectroscopy (Solar S2, Thermo Scientific, United States). Fe(II) determination
was carried out by the 1,10-phenanthroline method (Jeffery, Bassett, Mendham, & Denney,
1989) with an UV/VIS spectrophotometer (Lambda 25, PerkinElmer, United States). pH was
measured with a pH-meter (Crison basic 25 pH-meter). All the measurements were done
every 2 or 3 days.
3.1. Comparative study between mineral bioleaching and e-waste bioleaching
Based on previous studies (Benzal et al. 2016; Bosecker 1997; Dorado et al. 2012),
bioleaching of e-waste has been investigated at the same operational conditions than
chalcopyrite bioleaching: a solid-liquid ratio of 10 g/L with a particle size below 1 mm of
178
diameter, initial pH 2, 30 ºC of temperature, 120 rpm stirring and the same sample of
Acidithiobacillus ferrooxidans. As stated in sections 2.1 and 2.2, the chalcopyrite and the e-
waste used has 26.4 and 34.5% of copper content, respectively. Results of both processes
are shown in Fig. 1. It was observed higher copper recovery from electronic waste than from
chalcopyrite. In particular, e-waste bioleaching duplicated the copper recovered by ore
bioleaching. This could be attributed to the structure of the material bioleached because the
materials used are so different without a significant difference in concentration. It is also
observed that the e-waste allows copper recovering from the first day whereas the ore started
the recovering 3 days later. Another explanation for the behaviour observed is that copper in
e-waste is found as a metallic state in a simple structure in relation to chalcopyrite structure,
in which copper is found inside a complex crystalline structure.
Figure 33. Comparison between bioleaching of e-waste and bioleaching of chalcopyrite.
3.2. Effect of oxidation state of iron ion
In order to elucidate the role of iron in copper bioleaching from PCBs, experiments were
carried out in mineral medium solutions containing either Fe(III) ions or Fe(II) ions. The
experiments were performed in presence of microorganism (biotic) and without microorganism
(abiotic). Incubation was carried out for 23 days and samples were taken simultaneously
during the bioleaching process. The copper concentration was analysed and the recoveries of
copper extraction were determined for the four tests. Results are shown in figure 2. As it can
be seen, recoveries from biotic and abiotic experiments are very similar and no difference was
observed between experiments carried out with Fe(II) or Fe(III) ions. According to the
bioleaching mechanism described in the literature (equations 1 and 2 in the introduction),
copper oxidation is conducted by Fe(III) so, it would be expected that, in abiotic controls,
copper leaching occurred in the medium containing Fe(III) but not in the medium containing
Fe(II). Likewise, it would be expected that, in biotic tests, the introduction of the leaching agent
Fe(III) leaded to a faster copper leaching.
179
Figure 34. Copper recovery and pH evolution along time in the study of the ferrous and ferric
ions effect on bioleaching.
Regarding pH, from the initial pH 2, an increase up to 4 was observed in all the samples after
3 days. In particular, when ferric ions are involved, the increases of pH values were more
pronounced in comparison to the process using ferrous ions. Nevertheless the pH value
remained constant from the 13th day in the four experiments, nearly to values of 4.3. The
increase of pH observed the first days could be attributed to an alkalization produced by the
electronic waste itself. In this sense, PCB have been reported to be alkaline in nature (Wang,
Bai, Xu, & Liang, 2009). In order to a better understanding of the role that these two oxidation
forms of the iron have in the process, the concentration of Fe(II) and Fe(III) along experimental
time was analysed (Fig. 3). It is noteworthy, that regardless the oxidation state (Fe(II) or
Fe(III)) used to prepare the mineral medium, after one day the two media have similar
concentration of Fe(II), probably due to Fe(II) is easily oxidised by the air to Fe(III). In figure 3,
it is also noteworthy that the total iron is imperceptible after 13 days of experimentation. This
probably means that iron precipitated as hydroxide due to the alkalisation observed from the
first day. Precipitation of iron occurred when the pH is greater than 3 (Pourbaix, 1966).
Figure 35. Evolution of ferric ion, ferrous ion and total iron in abiotic samples using mineral
medium with ferrous ion (a) and with ferric ion (b): ferrous ion; ferric ion; and, total iron.
180
According to some authors, bacteria catalyses the transformation of Fe(II) to Fe(III), leaving
this last ion available to carry on with the bioleaching process. However, as it has stated above,
copper extraction was very similar in biotic and abiotic experiments. This means that chemical
leaching dominated the process and that microorganisms’ activity could have been masked
by some phenomena such as biosorption of copper onto the cellular wall of biomass or by the
inhibitory effect of some toxic materials from PCB. Our results agree with those of Lambert et
al. (2015) who also obtained the same copper recovery with microorganisms and without
them. They explained these results affirming that chemical mechanism of copper leaching is
faster than bio-oxidation. Nevertheless, they also explained that it is possible that bacteria
may have not been adapted to the test conditions and died during the experiment. PCB waste
contains different components, as plastics, that could be toxic for microorganisms as it has
stated previously, or even, the alkalinisation itself could be an inhibitory factor for the biological
activity. Under the experimental conditions in present study, the non-adaptation could be
associated to the amount of PCB treated (10 g/L). Brandl et al. (2001) concluded that
electronic waste alkalizes the mineral medium although this can be avoided if the
concentration used in the experiments not exceed 10 g/L. In the present study, the highest e-
waste concentration used was 10 g/L in order to avoid the alkalization of the medium in which
bioleaching took place, but despite of that, some alkalization was produced, reducing the
biological activity. This could be also the reason why in the present experiment the same
copper recovery from biotic and abiotic samples was obtained.
3.3. Biosorption test of copper by Acidithiobacillus ferrooxidans
Biosorption of metals by different microorganisms is a widely studied phenomenon (Fein,

2006; Javanbakht, Alavi, & Zilouei, 2014; Vijayaraghavan & Yun, 2008). In those cases
adsorption is a desirable effect since the later objective is water treatment. In particular, Ruiz-
Manríquez et al. (1997) studied the adsorption of copper onto Acidithiobacillus ferrooxidans,
concluding that an important biosorption was produced, however, they affirmed that the
phenomenon depends significantly on the pH. The present experiment was carried out at pH
= 2 and 30ºC to maintain the same conditions in which bioleaching process took place. The
results demonstrated that biosorption was produce at the conditions tested (Fig. 5). Freundlich
and Langmuir model isotherms were done to observe which one fitted better to experimental
data, being the first one the most accurate prediction. Nevertheless, the amount of copper
bioadsorpbed was not significant in comparison with the amount of copper extracted by
leaching. For this reason, the biosorption phenomenon can not be considered the responsible
for the similar results between biotic and abiotic experiments.
181
Figure 36. Biosorption of copper by Acidithiobacillus ferrooxidans.
3.4. Effect of e-waste concentration during bioleaching process
As it was showed above, concentrations of e-waste around 10 g/L could be not suitable for
bioleaching process. According to some authors (Adhapure et al., 2013; Xiang et al., 2010;
Yang et al., 2014; Zhu et al., 2011), the amount of initial copper could influence the process.
Xiang et al. 2010 studied the effect of e-waste dosage on bioleaching. They tested 10, 20, 30,
40 and 50 g/L of electronic waste, demonstrating that high PCB dosages resulted in lower
copper extraction. The same conclusion was obtained by the rest of the authors mentioned
above in this section, although all of them investigated with dosages over 4 g/L of PCB. Based
on this conclusion, the influence of the e-waste concentration from PCB on bioleaching was
studied by testing 2.5 and 5.0 g/L of PCB. Control essays were carried out without biomass
for all the dosages proved. Results of the copper recovery percentage along time are plotted
in figure 5.
Figure 37. Copper recovery along time at different electronic waste concentrations.
182
As it can be seen, the less amount of e-waste, the more recovery of copper was obtained. In
particular, a recovery close to 100% was achieved using microorganisms in the tests with 2.5
g/L of PCB. In this case, nearly twice the amount of copper retrieved by chemical leaching can
be obtained by means of the bioleaching. When 5.0 g/L of PCB were tested, the same
behaviour was observed, but differences between the abiotic and biotic samples were less
pronounced. However, no differences were observed between biotic and abiotic in 10 g/L test.
This means that biological activity is significantly affected by the waste dosage of e-waste,
which affect in turn the amount of copper bioleached. In the light of these results, this is
possible due to the e-waste contains some component that has an inhibitory effect on
Acidithiobacillus ferrooxidans growth and development, as Lambert et al. (2015) and Willner
(2013) observed in their studies.
Regarding the pH, as can be observed in figure 6, pH values increased from 2 to 4 the first 6
days and then, they remained quite constant for 5 days in all samples. Despite of this similar
behaviour between all samples, after 10 days of experimentation differences appeared within
them It was observed that the pH decreased to pH = 2 in the tests with 2.5 g/L of e-waste,
whereas the pH scant increased to pH = 4.5 in the tests with 10.0 g/L. In case of the tests with
5.0 g/L of PCB, they remained quite constant among pH 3 – 4. It is noteworthy that the increase
from 2 to 4 was more pronounced when the dosage of PCB increase. It confirms that PCB
alkalises the samples and the alkalinisation is bigger when there is more PCB in the solution.
This alkalinisation could considerably reduce the biological activity since it is well stablished
that Acidithiobacilus ferrooxidans growth and development take place at pH below 2.5
(Meruane & Vargas, 2003). These authors related the inhibition of the biological activity to the
formation of ferric oxide layer on bacteria, hindering the protons diffusion. For this reason we
investigated the bioleaching process under controlled pH adjusting pH values over
bioleaching process by addition of H2SO4 1:1 acid to keep pH = 2. Results obtained in those
experiments, carried out with and without pH control, are shown in figure 7.
Figure 38. Evolution of pH along time at different PCB dosages without pH control.
As it can be observed form figure 7, in both dosages tested (2.5 and 5.0 g/L of PCB,
respectively) copper recovering increases when the pH was adjusted; in particular, the
183
improvement was more noticeable during the first 10 days of experimentation. This could be
attributed to the fact that the adjustment of pH avoided the alkalization of the mineral medium
caused by the e-waste, as stated previously. This means that the biological activity was more
efficient than without pH control and so the efficiency of copper retrieved. However, after this
period, no significant improvement was observed when the pH was adjusted, independently
of the dosage treated.
Figure 39. Copper recovery along time in biotic samples with 2.5 and 5.0 g/L of PCB and with
pH control and without it.
4. CONCLUSIONS
When bioleaching of e-waste was compared with bioleaching of a mineral ore like chalcopyrite,
some differences were observed. The main one was the amount of copper extracted, obtaining
more copper bioleached using e-waste instead of chalcopyrite. This fact can be related to the
matrix in which copper is placed, because e-waste and the ore have quite different structures.
During the bioleaching of copper from electronic waste with a PCB concentration of 10 g/L, no
differences were obtained between biotic and abiotic tests. In case of adding Fe(III) ions
instead of Fe(II) ions in the mineral medium, more copper is extracted although the difference
was not too noticeable (around 10%). It was also observed that half of ferrous ion
concentration added was oxidized to ferric ion at the initial moment of the experiment. This is
probably the reason why the difference on using ferric or ferrous ion is not noticeable. The
similar copper leached in biotic and abiotic tests can be attributed to the possible toxicity of e-
waste to microorganisms using a ratio of 10 g/L of e-waste.
Moreover, the present study demonstrated that the microorganism Acidithiobacillus

ferrooxidans could biosorpt solubilized copper. However, the biosorption obtained at the pH 2
and 30 ºC (the conditions in bioleaching process) was not as significant to be considered
during the process.
Regarding to chemical and biological leaching at 2.5, 5.0 and 10.0 g/L e-waste concentrations,
it was observed that the concentration affected the process. In particular, it was demonstrated
that differences between biotic and abiotic tests are more pronounced when less e-waste
concentration is treated. This means that e-waste results toxic for microorganisms when the
184
amount of e-waste is too great because of its composition like plastics. Even so, the
improvement observed when Acidithiobacillus ferrooxidans are involved is noticeable,
obtaining up to 30% more copper recovery when 2.5 g/L of e-waste are treated biologically
instead of chemically. In case of controlling the pH of biotic samples of 2.5 and 5.0 g/L of PCB,
it was observed an improvement of copper bioleaching during the first 10 days of
experimentation because the adjustment avoided the alkalization produced by PCB. However,
after this period, there was not a significant improvement on copper recovery between
samples with adjusted pH and without it.
ACKNOWLEDGMENTS
Eva Benzal gratefully acknowledges a FPU-2014 predoctoral scholarship from Ministerio de

Educación, Cultura y Deporte (Spain).
5. REFERENCIES
Adhapure, N. N., Waghmare, S. S., Hamde, V. S., & Deshmukh, a. M. (2013). Metal
solubilization from powdered printed circuit boards by microbial consortium from bauxite
and pyrite ores. Applied Biochemistry and Microbiology, 49(3), 256–262.
http://doi.org/10.1134/S0003683813030034
Arshadi, M., & Mousavi, S. M. (2015). Multi-objective optimization of heavy metals bioleaching
from discarded mobile phone PCBs: Simultaneous Cu and Ni recovery using
Acidithiobacillus ferrooxidans. Separation and Purification Technology, 147, 210–219.
http://doi.org/10.1016/j.seppur.2015.04.020
Baldé, C. P., Wang, F., Kuehr, R., & Huisman, J. (2015). The global e-waste monitor - 2014.
Bonn, Germany: United Nations University, IAS-SCYCLE. Retrieved from
http://i.unu.edu/media/unu.edu/news/52624/UNU-1stGlobal-E-Waste-Monitor-2014-
small.pdf
Bas, a. D., Deveci, H., & Yazici, E. Y. (2013). Bioleaching of copper from low grade scrap TV
circuit boards using mesophilic bacteria. Hydrometallurgy, 138, 65–70.
http://doi.org/10.1016/j.hydromet.2013.06.015
Bosecker, K. (1997). Bioleaching: metal solubilization by microorganisms. FEMS Microbiology
Reviews, 20(3-4), 591–604. http://doi.org/10.1016/S0168-6445(97)00036-3
Brandl, H., Bosshard, R., & Wegmann, M. (2001). Computer-munching microbes: metal
leaching from electronic scrap by bacteria and fungi. Hydrometallurgy, 59(2-3), 319–326.
http://doi.org/10.1016/S0304-386X(00)00188-2
Chen, S., Yang, Y., Liu, C., Dong, F., & Liu, B. (2015). Column bioleaching copper and its
kinetics of waste printed circuit boards (WPCBs) by Acidithiobacillus ferrooxidans.
Chemosphere, 141, 162–8. http://doi.org/10.1016/j.chemosphere.2015.06.082
Choi, M.-S., Cho, K.-S., Kim, D.-S., & Kim, D.-J. (2004). Microbial recovery of copper from
printed circuit boards of waste computer by Acidithiobacillus ferrooxidans. Journal of
Environmental Science and Health. Part A, Toxic/hazardous Substances &
Environmental Engineering, 39, 2973–2982. http://doi.org/10.1081/LESA-200034763
Dorado, A. D., Solé, M., Lao, C., Alfonso, P., & Gamisans, X. (2012). Effect of pH and Fe(III)
ions on chalcopyrite bioleaching by an adapted consortium from biogas sweetening.
Minerals Engineering, 39, 36–38. http://doi.org/10.1016/j.mineng.2012.06.009
Fein, J. B. (2006). Thermodynamic Modeling of Metal Adsorption onto Bacterial Cell Walls:
Current Challenges. Advances in Agronomy, 90(06), 179–202.
http://doi.org/10.1016/S0065-2113(06)90005-4
Fornalczyk, A., Willner, J., Francuz, K., & Cebulski, J. (2013). E-waste as a source of valuable
metals. Archives of Materials Science and Engineering, 63(2), 87–92.
Javanbakht, V., Alavi, S. A., & Zilouei, H. (2014). Mechanisms of heavy metal removal using
microorganisms as biosorbent. Water Science and Technology, 69(9), 1775–1787.
185
http://doi.org/10.2166/wst.2013.718
Jeffery, G. H., Bassett, J., Mendham, J., & Denney, R. C. (1989). Vogel’s textbook of
quantitative chemical analysis (5th ed.). England: Longman Scientific & Technical.
Lambert, F., Gaydardzhiev, S., Léonard, G., Lewis, G., Bareel, P.-F., & Bastin, D. (2015).
Copper leaching from waste electric cables by biohydrometallurgy. Minerals Engineering,
76, 38–46. http://doi.org/10.1016/j.mineng.2014.12.029
Meruane, G., & Vargas, T. (2003). Bacterial oxidation of ferrous iron by Acidithiobacillus
ferrooxidans in the pH range 2.5-7.0. Hydrometallurgy, 71(1-2), 149–158.
http://doi.org/10.1016/S0304-386X(03)00151-8
Mražíková, A., Marcinčáková, R., Kaduková, J., & Velgosová, O. (2013). Influence of bacterial
culture to copper bioleaching from printed circuit boards. Inzynieria Mineralna, 14(2), 59–
62.
Pourbaix, M. (1966). Atlas of Electrochemical Equilibria in Aqueous Solutions. (Pergamon,
Ed.). New York.
Ruiz-Manríquez, A., Magaña, P. I., López, V., & Guzmán, R. (1997). Biosorption of Cu by
Thiobacillus ferrooxidans. Bioprocess Engineering, 18(2), 113–118.
http://doi.org/10.1007/s004490050420
Shah, M. B., Tipre, D. R., Purohit, M. S., & Dave, S. R. (2015). Development of two-step
process for enhanced biorecovery of Cu–Zn–Ni from computer printed circuit boards.
Journal of Bioscience and Bioengineering, 120(2), 167–173.
http://doi.org/10.1016/j.jbiosc.2014.12.013
Vijayaraghavan, K., & Yun, Y. S. (2008). Bacterial biosorbents and biosorption. Biotechnology
Advances, 26(3), 266–291. http://doi.org/10.1016/j.biotechadv.2008.02.002
Wang, J., Bai, J., Xu, J., & Liang, B. (2009). Bioleaching of metals from printed wire boards by
Acidithiobacillus ferrooxidans and Acidithiobacillus thiooxidans and their mixture. Journal
of Hazardous Materials, 172(2-3), 1100–1105.
http://doi.org/10.1016/j.jhazmat.2009.07.102
Willner, J. (2013). Influence of Physical and Chemical Factors on Biological Leaching Process
of Copper From Printed Circuit Boards. Metalurgija, 52(2), 189–192.
Willner, J., & Fornalczyk, A. (2013). Extraction of metals from electronic waste by bacterial
leaching. Environment Protection Engineering, 39(1), 197–208.
http://doi.org/10.5277/EPE130115
Xiang, Y., Wu, P., Zhu, N., Zhang, T., Liu, W., Wu, J., & Li, P. (2010). Bioleaching of copper
from waste printed circuit boards by bacterial consortium enriched from acid mine
drainage. Journal of Hazardous Materials, 184(1-3), 812–8.
Yang, Y., Chen, S., Li, S., Chen, M., Chen, H., & Liu, B. (2014). Bioleaching waste printed
circuit boards by Acidithiobacillus ferrooxidans and its kinetics aspect. Journal of
Biotechnology, 173, 24–30. http://doi.org/10.1016/j.jbiotec.2014.01.008
Zhu, N., Xiang, Y., Zhang, T., Wu, P., Dang, Z., Li, P., & Wu, J. (2011). Bioleaching of metal
concentrates of waste printed circuit boards by mixed culture of acidophilic bacteria.
Journal of Hazardous Materials, 192(2), 614–619.
186
Oral Presentations – Strategies Against Climate Change

187
Fluidized Bed Oxy-Combustion of Coal and

Biomass: an Expedient Way to Reduce Pollutant
Emissions
Carlos Lupiáñez1, Luis Ignacio Díez2
1CIRCE, Mariano Esquillor 15, 50018-Zaragoza, Spain. E-mail:carloslu@fcirce.es
2EINA, María de Luna 5, 50018-Zaragoza, Spain. E-mail: luisig@unizar.es
ABSTRACT
During last years, the development of bio-CCS technologies has emerged due to the
possibility of capturing neutral CO2 from biomass combustion. This increases the
attractiveness of CCS towards negative CO2 emissions. Among the CCS options, oxy-firing
has been developed to be now ready for commercial-scale facilities. Its application on fluidized
bed reactors additionally offers specific advantages, as low emission levels and use of low
rank fuels. The use of biomass in a combustion system implies several challenges related to
the wide variety of its composition and operational problems related to fouling and corrosion.
Co-firing coal and biomass has been found as a solution for several of the inconveniences,
also enabling the use of biomass in larger units. Aiming to characterize emissions from oxy-
firing of coal and biomass, an experimental campaign has been executed in a lab-scale
fluidized bed unit. Firstly, Spanish anthracite was fired with corn stover. Later, anthracite was
substituted by domestic lignite, with higher sulfur and ash contents. Coal-to-biomass share,
Ca:S ratio and chlorine content in the biomass were related to the emissions in the facility.
Combustion efficiency was always outstanding, even for the cases with anthracite. SO2
capture was affected by oxy-firing conditions, showing an increase in relation to the chlorine
content in the biomass. NOx emissions were slightly affected by the change of the biomass-
related parameters, being more dependent on the atmosphere or the operating conditions.
Keywords: Bio-CCS; Oxy-firing; Co-firing; Emissions.
1. INTRODUCTION
Fossil fuel combustion has been traditionally the main way to generate electricity. In spite of
the development of renewable sources and the improvement of energy efficiencies, global
CO2 emissions are continuously increasing. Facing this situation, the development of carbon
capture and storage (CCS) emerges as a reliable way to abate CO2 emissions in medium
term. The efforts carried out so far are close to get the purpose of commercially-available
units, and oxy-firing is one of the outstanding technologies (Santos, 2015). The development
of combustion in an O2/CO2 atmosphere was initially tested and applied to pulverized fuel
burners. The experiences on fluidized beds started later, but nowadays, the largest facilities
are in a similar size (Maier, 2015). These reactors offer some advantages, namely: firing low
rank fuels, with high ash and moisture contents, low NOx emissions due to the relatively low
combustion temperature, and in-situ desulphurization by means of the addition of a sorbent.
These aspects enhance the interest on the application of oxy-firing in fluidized bed reactors.
The use of residual biomass as fuel implies the emission of neutral CO2; then, the joint
combination of biomass and CCS leads to the concept of “negative” CO 2 emissions.
Nevertheless, biomass combustion gathers several issues related to its composition which
188
affect to ash deposition (fouling and slagging) and corrosion. A way to diminish these
affections is co-firing (Aho and Ferrer, 2005), which also facilitates the use of biomass in the
largest plants. Previous experiences of oxy-firing coal and biomass in fluidized bed are really
limited in literature. Tan et al. (2013) demonstrated the feasibility of co-firing coal and biomass
in fluidized bed under oxy-firing conditions. Duan et al. (2015) studied the influence of O2
concentration, oxidizer staging and bed temperature on NO emission, obtaining the same
trends described when oxy-firing coal. Kosowska-Golachowska (2015) reported a decrease
of burn-out time together with higher temperatures with the O2 enrichment of fluidizing gas.
Aiming to increase the knowledge about pollutant emission during oxy-firing coal and biomass,
this contribution analyses SO2 and NO emitted during oxy-firing anthracite and lignite together
with corn stover. The influence of parameters as fluidizing gas composition, calcium to sulphur
ratio, coal to biomass ratio and chlorine content in biomass are studied.
2.1. Facility
The experimental campaign has been executed in CIRCE oxy-firing laboratory (Figure 1). The
reactor is a bubbling fluidized bed (2.5 m height and 0.2 m inner diameter) which can be
operated under air-firing and oxy-firing conditions. For the former, air is taken from the ambient
by means of a driven fan, whereas for the latter a blend of O 2 and CO2 is taken from
commercial cylinders with a variable composition set in a mixer. Fuel is fed from two hoppers,
one for coal and sorbent and the other one for biomass. The material is discharged by two
variable-speed endless screw-feeders into another one which injects the fuel in the reactor. A
detailed description of the facility can be found elsewhere (Romeo et al., 2010).
Figure 40. Oxy-fired fluidized bed facility
189
Temperatures, pressures and flows are continuously monitored by sensors distributed along
the facility. Exhaust gas composition is determined by a gas analyser composed by three
modules. Two of them, based on non-dispersive infra-red technology (NDIR), measure CO,
CO2, NO and SO2, whereas the third one, with a paramagnetic sensor, gives O2 concentration.
2.2. Fuels and sorbent

Two Spanish coals were selected for the experimental campaign. The first one is an anthracite
from the North of Spain, with low moisture and high ash contents. The second one is a lignite
from Teruel (Spain) with a high sulphur content. As concerns biomass, an agricultural residue
as corn stover was selected. This biomass was chosen due to its chlorine content. Fuel
compositions are showed in Table 1.
Table 9. Proximate and ultimate analysis of the fuels

Anthracite Lignite Corn stover
Proximate analysis (%wt.)
Moisture 2.42 13.57 6.18
Ash 31.30 30.30 5.50
Volatiles 6.71 25.72 70.68
Fixed carbon 59.57 30.41 17.64
Ultimate analysis (%wt.)
C 59.27 40.53 43.3
H 2.01 3.18 5.82
N 0.91 0.28 0.57
S 1.58 6.65 0.11
Cl – – 0.35
LHV (kJ/kg) 21 620 14 434 15 438
As sorbent, Granicarb limestone from Belchite (Spain) was used, which is characterized by its
high CaCO3 content (>99%). Particle size ranged 0.1-1.0 mm and the mean particle diameter
was 0.6 mm. Initial bed was composed by 5 kg of silica sand with a similar mean diameter
than that of limestone.
2.3. Experimental matrix

Ten experiments were executed, five with each coal, blended with corn stover (see Table 2).
The experimental matrix was design to allow the discussion of the influence of the most
relevant paremeters.
Table 10. Experimental conditions of the experiments

Coal to Cl
Test # Fuel Fluidizing gas Ca:S
biomass ratio (%wt.)
A1 Anthracite Air 80/20 0.0 1.00
A2 Lignite Air 80/20 2.5 1.00
OXY1 Anthracite 30/70% O2/CO2 80/20 0.0 1.00
OXY5 Lignite 30/70% O2/CO2 80/20 2.5 1.00
OXY6 Lignite 30/70% O2/CO2 90/10 2.5 1.00
OXY7 Lignite 30/70% O2/CO2 80/20 4.0 0.35
OXY8 Lignite 30/70% O2/CO2 80/20 4.0 1.00
Air-firing and oxy-firing (30/70% O2/CO2) tests were addressed to compare the influence of
combustion atmosphere. To study the effect of fuel composition, two coal to biomass ratios,
190
80/20% and 90/10% in energy basis, were fed. The Ca:S ratios injected during the anthracite
experiments were 0 and 2.5. This allows to study the influence of the presence of limestone.
Nevertheless, when feeding lignite, the absence of limestone is not recommended due to the
high sulphur content, so two Ca:S was chosen, 2.5 and 4. Two levels of chlorine in biomass
were tested (0.35% and 1%). The first one is that contained in the original corn stover, whereas
the other one was obtained by means of doping the biomass with a KCl solution.
The initial silica sand bed is preheated by a propane burner up to a temperature ranging 500-
600ºC, depending on the coal. Then, burner is shut down and coal feeding starts under air-
firing conditions up to 850ºC. In the case of air-firing, coal and biomass feeding rates are
adjusted to coal/biomass ratio desired. If experiment is executed under oxy-firing conditions,
combustion mode is changed at 850ºC and then operation becomes similar than in air-firing
case. When steady state conditions are reached, operational data are collected during one
hour and a half for further analysis.
Table 11 shows the temperatures measured during the experiments together with the pollutant
emissions registered by the gas analyser. The comparison of the pollutant concentrations of
the air-firing and oxy-firing tests is affected by the different volume of fluidizing gas. So, fuel-
N to NO conversion and desulphurization are also calculated to allow a proper comparison.
Table 11. Bed temperature, emissions (normalized to 6% O2) and calculated desulphurization
efficiency in exhaust gases
Tbed CO2 CO NO NO/Fuel-N SO2 Desulphurization

O2 (%)
(ºC) (%) (mg/Nm3) (mg/Nm3) (%) (mg/Nm3) efficiency (%)
A1 857 5.19 15.49 1007 184 7.73 1822 42.26
A2 876 5.66 14.35 1139 240 20.1 2207 87.71
OXY1 850 8.04 91.78 669 626 13.8 2671 57.41
OXY2 865 4.99 92.81 457 595 15.09 3101 46.99
OXY3 847 13.40 86.49 788 843 21.9 590 90.65
OXY4 870 11.76 88.20 N/A 648 16.94 168 96.56
OXY5 856 5.85 93.62 746 343 16.09 12155 61.66
OXY6 852 3.42 95.99 473 514 27.71 13493 59.51
OXY7 851 1.61 97.91 417 514 24.33 9790 70.16
OXY8 862 2.31 97.22 908 289 14.69 8671 71.45
3.1. NO
NO emission is affected by the combustion mode according to the results obtained. In the
case of anthracite, the expected increase of NO concentration in exhaust gases as a
consequence of the different gas-flows was measured. The same behaviour was detected for
fuel-N conversion to NO. On the other hand, burning lignite and corn stover resulted in higher
NO concentration in oxy-firing but lower fuel-N conversion. Other parameters as excess O2 or
thermal input has a high influence on NO and the effect of changing combustion mode can be
overlapped. Previous experiences reported lower fuel-N conversion under O2/CO2
atmospheres due to the promotion of gasification reactions and the formation of CO (Duan et
al., 2011). This has the capability of reducing NO by direct reaction or by catalysing the NO
destruction reaction on the char surface.
191
The increase of anthracite to biomass ratio in OXY2 resulted in similar NO emission than that
from OXY1, but with a slightly higher fuel-N conversion. This can be expected due to the higher
N in anthracite. Tests OXY5 and OXY6 (with lignite) showed an increase of NO emitted and
fuel-N conversion when increasing the coal in the fuel blend. NO diminution should be
expected in OXY6 due to the reduction of volatile matter and fuel-N injected, together with the
lower excess O2. So, the result suggests the suppression of some NO reducing reactions with
coal ratio increase that enhances a higher NO emission.
No relevant influence was found due to the increase of Ca:S when comparing tests OXY1
against OXY4 and OXY5 against OXY8. Several studies carried out under conventional
combustion conditions reported an increase of NO emitted when limestone was added (de
Diego et al., 1996). Sorbent addition enhances NO formation by modifying several NO
formation mechanisms (Liu et al., 2001). Nevertheless, the absence of calcination in oxy-firing
limits the effect of sorbent on NO due to the low catalysing effect of CaCO 3 (Lupiáñez et al.,
2013).
The chlorine content in biomass had a negligible effect on NO emission in both combustion
modes. The variation of NO observed when chlorine content was increased is related to the
difference in CO or in excess O2.
3.2. SO2
The influence of fluidizing gas composition can be studied by comparison of test A1 against
OXY1 (anthracite) and A2 against OXY5 (lignite). For the anthracite experiments, the change
of combustion mode to oxy-firing implied a small increase of SO2 concentration, but regarding
the values of desulphurization efficiency, it is more than 20% higher in the case of OXY1 than
in A1. The opposite behaviour is observed in the case of lignite experiments. SO2
concentration was clearly higher in OXY5 and desulphurization efficiency is 25% lower than
in A2.
The main cause of this different behaviour is related to the SO2 capture mechanism. The
absence of limestone in anthracite experiments implied that SO2 retention is a consequence
of auto-retention, i.e. SO2 is retained in the ashes generated during combustion that remains
in the reactor. According to the results, this mechanism is favoured in oxy-firing conditions.
In the case of lignite experiments, the difference is a consequence of the change of

desulphurization mechanism. Under air-firing conditions, the low CO2 partial pressure together
with the bed temperature allows CO2 release, that is, limestone calcination (R.1). The result of
this reaction is CaO, which is characterized by having larger pores than limestone which
facilitates the diffusion of SO2 in the particle. Then, CaO is sulphated generating CaSO4 (R.2):
CaCO3 ↔ CaO + CO2 R.1
CaO + SO2 + ½ O2 → CaSO4 R.2
In oxy-firing, the typical high CO2 partial pressure inhibits limestone calcination at 850ºC, so
desulphurization is carried out by direct mechanism (R.3):
192
CaCO3 + SO2 + ½ O2 → CaSO4 + CO2 R.3
The smaller pores of limestone decrease the use Ca available in a particle due to pore blocking
when CaSO4 is formed. Therefore, direct desulphurization results in a lower desulphurization
efficiency (de Diego et al., 2011).
The effect of increasing coal to biomass ratio gave the expected result, that is, an increase of
SO2 emitted. The main source of sulphur in fuel was S-coal, so the higher the coal feeding
rate, the higher SO2 in exhaust gases.
Increasing Ca:S gave a reduction of SO2 concentration. Nevertheless, the effect is more
relevant in the experiments with anthracite due to the initial Ca:S was 0. In fact, the
desulphurization efficiency in test OXY4 was over 95% under direct capture conditions.
According to this result, direct capture conditions with a Ca:S of 2.5 can be considered enough
for this blend of fuel from the point of view of SO2 emission. When lignite and corn stover were
burnt, desulphurization also improves with higher Ca:S, but the efficiency remains clearly
lower than in the case of air-firing.
The augmentation of chlorine content in biomass yielded higher desulphurization efficiencies.

This result was found when facing tests OXY3 against OXY4 and OXY7 against OXY8.
Chlorine is known by its high volatility and its capability to form alkali chlorides in gaseous
phase. This enhances the retention of sulphur by means of reaction with alkali metal according
the following mechanisms (R.4 and R.5):
2 MCl + SO2 + ½ O2 + H2O  M2SO4 + 2 HCl R.4
2 MCl + SO3 + H2O  M2SO4 + 2 HCl R.5
Where M is the alkali metal.
The analysis of the deposits confirmed the higher presence of alkali sulphates in the
experiments with the highest chlorine content.
4. CONCLUSIONS
Oxy-firing of different blends of coal and biomass is an expedient way to reduce emissions
from combustion systems, increasing the benefits of bio-CCS technologies. Fluidized bed
reactors are able to further offer the flexibility of firing very different coals, as the reported in
this work (anthracite and lignite). NO emissions are reduced under oxy-firing, but parameters
as excess O2 or CO concentration have a strong influence on the results. The absence of
calcination, due to the high CO2 concentration, inhibits the catalysing effect of increasing Ca:S,
whereas no relation has been found between chlorine content in biomass and NO. Shifting
combustion mode to oxy-firing can improve desulphurization by means of auto-retention, but
desulphurization worsens at typical fluidized bed temperatures. The latter can be balanced by
the augmentation of Ca:S ratio. SO2 capture efficiency is enhanced with high chlorine content
in biomass and decreasing the coal ratio in the fuel blend.
193
ACKNOWLEDGMENTS
The work described in this paper was partially funded by the R+D Spanish National Program
from the Spanish Ministry of Economy and Competitiveness, under the Project ENE2012-
39114. The project is also co-funded by the European Commission (European Regional
Development Funds).
REFERENCES
Aho M, Ferrer E (2005) Importance of coal ash composition in protecting the boiler against chlorine
deposition during combustion of chlorine-rich biomass. Fuel 84:201–212.
de Diego LF, Londono CA, Wang XS, Gibbs BM (1996) Influence of operating parameters on NOx and
N2O axial profiles in a circulating fluidized bed combustor. Fuel 75:971–978.
de Diego LF, de las Obras-Loscertales M, García-Labiano F, Rufas A, Abad A, Gayán P, Adánez J
(2011) Characterization of a limestone in a batch fluidized bed reactor for sulfur retention under
oxy-fuel operating conditions. International Journal of Greenhouse Gas Control 5:1190–1198.
Duan L, Zhao C, Zhou W, Qu C, Chen X (2011) Effects of operation parameters on NO emission in an
oxy-fired CFB combustor. Fuel Processing Technology 92:379–384.
Duan L, Duan Y, Zhao C, Anthony EJ (2015) NO emission during co-firing coal and biomass in an oxy-
fuel circulating fluidized bed combustor. Fuel 150:8–13.
Kosowska-Golachowska M, Wolski K, Kijo-Kleczkowska A, Musiał T, Środa K (2015) Experimental
Research on Oxy-Fuel Combustion of Biomass in a Circulating Fluidized-Bed. 7th European
Combustion Meeting (ECM 2015), March 30-April 2. Budapest, Hungary.
Liu H, Gibbs BM (2001) The influence of calcined limestone on NOx and N 2O emissions from char
combustion in fluidized bed combustors. Fuel 80:1211–1215.
Lupiáñez C, Díez LI, Romeo LM (2013) NO Emissions from Anthracite Oxy-Firing in a Fluidized-Bed
Combustor: Effect of the Temperature, Limestone, and O2. Energy & Fuels 27:7619–7627.
Maier J (2015) Introduction and overview of oxyfuel combustion for power plant with CCS. Oxyfuel
Combustion Capacity Course, October 27. Wuhan, China.
Romeo LM, Díez LI, Guedea I, Bolea I, Lupiáñez C, González A, Pallarés J, Teruel E (2010) Design
and operation assessment of an oxyfuel fluidized bed combustor. Experimental Thermal and
Fluid Science 35:477–484.
Santos S (2015) Future of oxyfuel combustion. 5th Oxyfuel Combustion Research Network Meeting,
October 28-30. Wuhan, China.
Tan Y, Jia L, Wu Y (2013) Some Combustion Characteristics of Biomass and Coal Cofiring under Oxy-
Fuel Conditions in a Pilot-Scale Circulating Fluidized Combustor. Energy & Fuels 27:7000–
7007.
194
Lanthanum modified hydrotalcites for CO2 capture

in sorption enhanced water gas shift (sewgs)
processes
Yarima Torreiro1, Marta Maroño1, Juan García2, José María Sánchez1
1 CIEMAT. División de Combustión y Gasificación. Avenida Complutense, 40. 28045, Madrid,

(Spain). yarima.torreiro@ciemat.es; marta.marono@ciemat.es; josemaria.sanchez@ciemat.es
2 Grupo de Catálisis y Procesos de Separación, Departamento de Ingeniería Química, Facultad
Ciencias Químicas, Universidad Complutense de Madrid. 28040, Madrid, (Spain).
juangcia@quim.ucm.es
ABSTRACT
Based on previous research conducted by CIEMAT, the commercial hydrotalcite MG61 (Mg/Al
= 2 and promoted with 20 wt% of K2CO3) from SASOL proved to have good CO2 capture
capacity under water-gas–shift (WGS) conditions (i.e, temperatures in the range of 300-500ºC
and presence of steam). In order to determine the effect on capture capacity of adding a
second promoter, the K-promoted MG61 sorbent was impregnated with lanthanum. For
comparison purposes lanthanum was also added to the unpromoted sorbent. Samples of
promoted and unpromoted hydrotalcite were calcined at 600ºC and impregnated with
lanthanum (5 wt%, 15 wt% and 25 wt%) by incipient wetness impregnation. Then, CO2 capture
tests for a binary mixture (15% CO2/85% N2) were carried out in a thermobalance at
atmospheric pressure and in the temperature range of 200 to 600ºC, in order to investigate
the performance of the different materials. In this work, the results obtained in CO2 capture
tests, as well as the effect of the addition of lanthanum in the preselected sorbent, promoted
and unpromoted, will be presented and discussed.
Keywords: K promoted hydrotalcite; CO2 capture; SEWGS reaction; upgrading of biomass
gasification syngas; lanthanum.
1. INTRODUCTION
Depending on the operating temperature, several materials (e.g. activated carbons, zeolites,
metal–organic frameworks, hydrotalcites, etc.) are used to capture CO2 by adsorption (Miguel
et al., 2014) . Among them, hydrotalcites can be used at intermediate temperatures (473–
673 K), which is the temperature range of interest for CO2 capture in sorption-enhanced water
gas shift SEWGS reaction processes. SEWGS is one of the most promising alternatives to
the use of liquid amines and PSA for H2 production and CO2 capture in pre-combustion
processes (M. Maroño et al., 2014). Hydrotalcite (HT) materials are layered double
hydroxides, which general formula is:
[M2+1-X M3+X (OH)2][An-] x/n y H2O (1)
where M2+ and M3+ are divalent (e.g. Mg2+, Ni2+, Zn2+, etc.) and trivalent (e.g. Al3+, Ga3+, Mn3+,
etc.) metal cations, respectively. An− is a charge balancing anion (CO32−, Cl−, SO42−, etc.)
located in the interlayer space and x is generally between 0.17 and 0.33 (Cavani et al., 1991;
Reichle, 1986), while y is the number of moles of hydration water located also in the interlayer
space. The mineral known as hydrotalcite has Mg2+, Al3+ as divalent and trivalent cations and
CO32− as balancing anion, respectively. Hydrotalcites can be found in nature or be
synthesized. Currently, several thousands of tons of hydrotalcites are yearly produced by
195
several chemicals companies (e.g. BASF, SASOL, Clariant, Kisuma Chemicals, Sakai
Chemicals, etc.) which found application as catalysts, precursors, catalyst support,
adsorbents, polymer stabilizers and antacids (Miguel et al., 2014).
In recent years, hydrotalcites and especially the mixed oxides formed by their reversible
thermal decomposition are becoming increasingly important for catalytic reactions. The main
advantage of the hydrotalcites as catalysts comes from the opportunity to control to a certain
degree their acid-basic properties (Bîrjega et al., 2005). Loading of several elements as
vanadium (Meira et al., 2006), gallium (Miguel et al., 2014; Yavuz et al., 2009), zirconium (Velu
et al., 1998), cobalt (Li et al., 2009), lithium (Shumaker et al., 2007), zinc (Rahul et al., 2011),
chrome, iron (Macala et al., 2008), nickel (Zhang et al., 2010) and rare earth elements (Bîrjega
et al., 2005) and how their addition enhanced the basicity and thereby the catalytic activity of
hydrotalcites have been studied.
Concerning the natural mineral Mg-Al-CO3, many works about the influence of several
parameters on their high temperature CO2 capture have been reported. For example, the
effects of the Mg/Al ratio (Di Cosimo et al., 1998) (Sharma et al., 2008), the thermal pre-
treatment (Sharma et al., 2008), the alkali doping (Lee et al., 2007; Oliveira et al., 2008), the
effects of divalent and trivalent cations and charge compensating anions (Pergher et al., 1999;
Sharma et al., 2008) or the influence of the synthesis method (Di Cosimo et al., 1998; Sharma
et al., 2008).
The deposition of metal cations such as alkaline or alkaline earth metals on MgO-based
catalysts for producing new centers with different acid-base properties is widely reported in
the literature (Choudhary et al., 1991; Díez et al., 2006; Ferretti et al., 2011; Matsuhashi et al.,
2000; Tang et al., 2011). The improvement in MgO catalytic activity in several reactions has
also been proved when it is impregnated with rare earth metals (Choudhary et al., 1997) and
specifically with lanthanum (Choudhary et al., 1989; Choudhary et al., 2000). An improvement
of CO2 capture capacity of alumina has also been reported when impregnation with alkali,
alkali earth metals or rare earth elements was carried out (Horiuchi et al., 1998).
The basicity of a surface oxide is generally related to the electrodonating properties of the
combined oxygen anions, so that the higher the partial negative charge on the combined
oxygen anions, the more basic the oxide. That is the reason why the promotion of MgO and
alumina (and hence, of calcined hydrotalcites) with more basic oxides (i.e., oxides with cations
with lower electronegativity than Mg and Al and thereby with a higher difference respect
oxygen´s electronegativity) has proved to increase their basicity (Dıe
́ z et al., 2000; Horiuchi et
al., 1998; Matsuhashi et al., 2000).
With regard to hydrotalcites, incorporated cations can have a higher ionic radius than those
present in the structure or a similar one (Rahul et al., 2011). In the first case, cations would
most likely reside in the interlayer space of HT as a separate oxide/hydroxide phase or on the
external surface or HT crystallites or as independent segregated phases (Shumaker et al.,
2007) (Li et al., 2009; Macala et al., 2008) (Miguel et al., 2014; Wang et al., 2010; Yavuz et
al., 2009). In the second case, elements added would be present in the framework of HT
(Bîrjega et al., 2005; Miguel et al., 2014; Wang et al., 2010).
Promotion of HTs with cations with a larger ionic radius than Mg+2 and Al3+ (Yavuz et al., 2009)
and with a similar ionic radius (Miguel et al. 2014) has been reported to enhance their CO2
capture capacity. The intercalation of long carbon-chain organic anions have also been proved
to markedly increase the CO2 capture capacity of these materials (Hutson et al., 2008; Roelofs
et al., 2002; Wang et al., 2010; Wang et al., 2012).
196
Lanthanum´s high ionic radius prevents the easy intercalation of La3+ in the hydrotalcites
galleries (Bîrjega et al., 2005; Li et al., 2009; Rahul et al., 2011; Zhang et al., 2010). When this
element is placed in the interlayer space the increase of the partial negative charge on the
combined oxygen anions can be favoured due to its strong anionic character and therefore,
the basicity of HTs (Bîrjega et al., 2005) (Li et al., 2009; Rahul et al., 2011). Table 1 shows the
ionic radius and electronegativity values of main elements present in this work.
Table 1. Ionic radius and electronegativity values of elements studied in this work
Electronegativity Ionic Radius (Å)
Mg 1.31 0.65
Al 1.61 0.50
La 1.10 1.15
K 0.82 1.33
O 3.44 1.40
Nevertheless the incorporation of La3+ species to hydrotalcites had thrown some doubts about
the place of its incorporation. Other authors claimed that if lanthanum contents are lower than
6.5 mol.%, this element could be introduced into the HT structure effectively (Zhang et al.,
2010).
Based on the discussion above, the aim of this work is to assess if the addition of lanthanum
as second promoter enhances CO2 capture capacity of K-promoted hydrotalcites under
SEWGS conditions. For comparison purposes, the effect of adding such a promoter in the
CO2 capture capacity of the unpromoted sorbent is also studied.

2.1. Materials
A Mg-Al hydrotalcite promoted with potassium carbonate (HT-K2CO3), which in a previous

paper (Marta Maroño et al., 2014) have shown to be appropriate for CO2 capture in terms of
sorption capacity under WGS conditions and regenerability was used for the experiments
presented in this work. Mg/Al ratio in K-promoted hydrotalcite is 2 and it was supplied
impregnated with 20 wt% K2CO3. The unpromoted hydrotalcite (HT) was also studied for
comparison purposes. Both materials were supplied by SASOL in pellets of 5x5 mm.
2.2. Sorbent preparation
Promoted and unpromoted hydrotalcite were firstly calcined in static air at 600ºC for 4 h. A
series of lanthanum (5-15-25 wt%) containing samples were prepared from both promoted
and unpromoted calcined hydrotalcites by the incipent wetness impregnation method.
Lanthanum nitrate was dissolved in deionized water and a predetermined volume was added
drop wise to the HT until it appeared wet. The wet powder was dried at 110ºC for 24h and
samples were then calcined in static air at 600ºC for 4 h.
2.3. CO2 capture tests
In order to investigate the performance of the different materials in the range of temperatures
of the WGS reaction (i.e, 300-500ºC), a series of CO2 capture tests were carried out in a
thermobalance at atmospheric pressure. About 50 mg of sample was placed in a platinum
basket, heated in N2 (100 ml/min) up to 600ºC at a heating rate of 10ºC/min and held at that
temperature for one hour. Then, the heater was shut off and the sample was cooled down to
197
200ºC using a flow of 100ml/min of CO2/N2 (15% CO2). The capture of CO2 is identified in each
sample by its increase in mass measured during the tests.

The effect of lanthanum promotion was investigated by measuring the weight gain of promoted
and unpromoted hydrotalcites impregnated with different weight percentages. Figure 1 shows
the results of the CO2 dynamic capture tests for all the sorbents prepared. As mentioned
above, in these experiments the capture of CO2 is identified by the increase in mass measured
during the tests.
Figure 1. Increase in mass due to CO2 capture during cooling of all samples from 600ºC to
200ºC at atmospheric pressure. Feed gas 15% CO2/N2.
As can be observed in Figure 1, potassium promoted hydrotalcite (HT-K2CO3) calcined at

600ºC is the sorbent which had the highest increase in mass (due to CO2 capture) in the
temperature range studied, as showed by the slope of the mass increase curve. Compared to
HT-K2CO3, the HT-K2CO3-La samples showed a significant decrease in CO2 capture capacity
(≈20% when HT-K2CO3 was impregnated with 5-15wt% of lanthanum and about 40% when
the percentage of added lanthanum was 25 wt%). The higher the amount of lanthanum added,
the lower the mass increase showed by the HT-K2CO3 sample. Nevertheless in comparison
with all the unpromoted HT samples the HT-K2CO3-La samples still caused an increase in
mass and, therefore, in CO2 capture capacity.
Some authors claim that K+, due to its ionic radius (1.33 Å), resides in the galleries between
the cationic layers, that means, in the interlayer space (Miguel et al., 2014; Yavuz et al., 2009).
However, other authors suggest that K+ could substitute Mg2+ in the HT framework and that
198
additional oxygen vacancies at the surface are created. CO2 would be adsorbed by oxygen
atoms located in the vicinity of K+(Meis et al., 2010). What is true is that values of K2CO3 up
to 22 wt% have been proved to drastically diminish the CO2 capture of hydrotalcites probably
due to pore or surface blockage (Walspurger et al., 2008). The potassium promoted
hydrotalcite studied in this work is a commercial material synthetized with 20% of K 2CO3. As
lanthanum ionic radius (1.15 Å) is close to that showed by potassium (1.33 Å) it is likely that
lanthanum may occupy the same place as potassium does in hydrotalcites.
If both promoters are placed in the same HT´s region and it has been proved that there is a
critical weight percentage to be added of promoters such as potassium, this would explain that
the higher the lanthanum content in K-promoted samples, the lower the CO2 capture capacity
probably due to a dominant blocking effect. Regarding the unpromoted hydrotalcite calcined
at 600ºC, results obtained are completely different and samples’ behaviour seemed to be
really dependent on temperature range studied.
At higher temperatures (300-600ºC), which is the temperature range of interest in SEWGS

processes, the addition of La improves CO2 capture, being sample HT-25%La the one with
the highest mass increase. As previously mentioned, adding a promoter in an unpromoted
hydrotalcite enhances its CO2 capture capacity. However, the second sample which showed
the highest mass increase was HT without lanthanum, followed by HT-5%La and HT-15% La,
which suggested us that there should be an optimum value of promoter to be added to
enhance HT CO2 capture capacity in the temperature range studied. However, at
temperatures close to 300ºC thermogravimetric profiles of samples HT-25% La and HT
intersect and, from that point and at temperatures below 275ºC (200-275ºC), samples HT and
HT-5% La are the samples which showed the highest mass increase (and hence, CO2 uptake).
4. CONCLUSIONS
A commercial hydrotalcite previously studied by CIEMAT and which proved to have good CO2
capture capacity under water-gas-shift (WGS) conditions was impregnated with lanthanum in
order to determine the effect of adding a second promoter in its CO2 capture capacity. For
comparison purposes lanthanum was also added to the unpromoted sorbent. Samples of
promoted and unpromoted hydrotalcite were calcined at 600ºC and impregnated with
lanthanum (5 wt%, 15 wt% and 25 wt%) by incipient wetness impregnation. Then, CO2 capture
tests for a binary mixture (15% CO2/85% N2) were carried out in a thermobalance at
atmospheric pressure and in the temperature range of 200 to 600ºC, in order to investigate
the performance of the different materials. From results obtained, the following conclusions
can be drawn:
 Potassium promoted hydrotalcite (HT-K2CO3) calcined at 600ºC is the sorbent which

showed the highest CO2 capture in the temperature range studied (200-600ºC).
 The higher the amount of lanthanum added in HT-K2CO3, the lower the mass increase
showed by the sample. Results might indicate that K and La, due to its ionic radius,
occupies the same region in hydrotalcites and that there is a critical value in the content
of promoters added from which a dominant blocking effect overcomes the increase in
basicity caused by promoters addition.
 Nevertheless and compared to all the unpromoted HT the HT-K2CO3-La samples still
caused an increase in mass and, hence, in CO2 uptake in the whole temperature range
studied.
 Regarding the unpromoted hydrotalcite (HT) calcined at 600ºC samples behavior
seemed to be really dependent on temperature.
199
o At temperatures between 300ºC-600ºC the increase of lanthanum content enhances

CO2 capture, being sample HT-25%La the one with the highest mass increase directly
followed by HT without lanthanum. Adding La as a promoter to the unpromoted
hydrotalcite improves its CO2 capture capacity but there seems to be an optimum value
of promoter to be added to enhance HT CO2 capture capacity.
o At temperatures below 275ºC thermogravimetric profiles of samples HT-5% La and HT
are those who showed de highest mass increase.
Unfortunately, from the results obtained it seems that the addition of lanthanum as a second
promoter does not improve the CO2 capture capacity obtained with potassium promoted
hydrotalcites. To fully understand the adsorption behaviour presented by samples as well as
the role of lanthanum further research as samples characterization by XRD and SEM is being
carried out.
ACKNOWLEDGMENTS
Authors wish to thank the Spanish Ministry of Economy (INNPACTO BIOH2 Project, IPT-2012-
0365-120000) and the Regional Government of Madrid provided through Projects
RETOPROSOST P2013/MAE-2907, REMTAVARES S2013/MAE-2716 and the European
Social Fund for their financial support.
REFERENCES
Bîrjega, R., Pavel, O. D., Costentin, G., Che, M., & Angelescu, E. (2005). Rare-earth elements modified
hydrotalcites and corresponding mesoporous mixed oxides as basic solid catalysts. Applied
Catalysis A: General, 288(1–2), 185-193.
Cavani, F., Trifirò, F., & Vaccari, A. (1991). Hydrotalcite-type anionic clays: Preparation, properties and
applications. Catalysis Today, 11(2), 173-301.
Choudhary, V. R., Chaudhari, S. T., & Pandit, M. Y. (1991). Highly stable Li-doped magnesium oxide
catalyst for oxidative conversion of methane to higher hydrocarbons. Journal of the Chemical
Society, Chemical Communications(17), 1158-1159.
Choudhary, V. R., Chaudhari, S. T., Rajput, A. M., & Rane, V. H. (1989). Oxidative coupling of methane
to C2-hydrocarbons over la-promoted MgO catalysts. Journal of the Chemical Society,
Chemical Communications(9), 555-556.
Choudhary, V. R., Rane, V. H., & Chaudhari, S. T. (1997). Surface properties of rare earth promoted
MgO catalysts and their catalytic activity/selectivity in oxidative coupling of methane. Applied
Catalysis A: General, 158(1–2), 121-136.
Choudhary, V. R., Rane, V. H., & Chaudhari, S. T. (2000). Factors influencing activity/selectivity of La-
promoted MgO catalyst prepared from La- and Mg- acetates for oxidative coupling of methane.
Fuel, 79(12), 1487-1491.
Di Cosimo, J., Dıez, V., Xu, M., Iglesia, E., & Apesteguıa, C. (1998). Structure and surface and catalytic
properties of Mg-Al basic oxides. Journal of Catalysis, 178(2), 499-510.
Díez, V. K., Apesteguía, C. R., & Di Cosimo, J. I. (2006). Aldol condensation of citral with acetone on
MgO and alkali-promoted MgO catalysts. Journal of Catalysis, 240(2), 235-244.
Dı́ez, V. K., Apesteguı́a, C. R., & Di Cosimo, J. I. (2000). Acid–base properties and active site
requirements for elimination reactions on alkali-promoted MgO catalysts. Catalysis Today,
63(1), 53-62.
Ferretti, C. A., Apesteguía, C. R., & Di Cosimo, J. I. (2011). MgO-based catalysts for monoglyceride
synthesis from methyl oleate and glycerol: Effect of Li promotion. Applied Catalysis A: General,
399(1–2), 146-153.
Horiuchi, T., Hidaka, H., Fukui, T., Kubo, Y., Horio, M., Suzuki, K., & Mori, T. (1998). Effect of added
basic metal oxides on CO2 adsorption on alumina at elevated temperatures. Applied Catalysis
A: General, 167(2), 195-202.
Hutson, N., & Attwood, B. (2008). High temperature adsorption of CO 2 on various hydrotalcite-like
compounds. Adsorption, 14(6), 781-789.
Lee, K. B., Verdooren, A., Caram, H. S., & Sircar, S. (2007). Chemisorption of carbon dioxide on
potassium-carbonate-promoted hydrotalcite. Journal of Colloid and Interface Science, 308(1),
30-39.
200
Li, E., Xu, Z. P., & Rudolph, V. (2009). MgCoAl–LDH derived heterogeneous catalysts for the ethanol
transesterification of canola oil to biodiesel. Applied Catalysis B: Environmental, 88(1–2), 42-
49.
Macala, G. S., Robertson, A. W., Johnson, C. L., Day, Z. B., Lewis, R. S., White, M. G., . . . Ford, P. C.
(2008). Transesterification catalysts from iron doped hydrotalcite-like precursors: solid bases
for biodiesel production. Catalysis Letters, 122(3-4), 205-209.
Maroño, M., Barreiro, M. M., Torreiro, Y., & Sánchez, J. M. (2014). Performance of a hybrid system
sorbent–catalyst–membrane for CO2 capture and H2 production under pre-combustion
operating conditions. Catalysis Today, 236, Part A, 77-85.
Maroño, M., Torreiro, Y., Montenegro, L., & Sánchez, J. (2014). Lab-scale tests of different materials
for the selection of suitable sorbents for CO 2 capture with H2 production in IGCC processes.
Fuel, 116, 861-870.
Matsuhashi, H., Oikawa, M., & Arata, K. (2000). Formation of Superbase Sites on Alkaline Earth Metal
Oxides by Doping of Alkali Metals. Langmuir, 16(21), 8201-8205.
Meira, D. M., Cortez, G. G., Monteiro, W. R., & Rodrigues, J. A. J. (2006). Vanadium oxides supported
on hydrotalcite-type precursors: the effect of acid-base properties on the oxidation of
isopropanol. Brazilian Journal of Chemical Engineering, 23, 351-358.
Meis, N. N. A. H., Bitter, J. H., & de Jong, K. P. (2010). On the Influence and Role of Alkali Metals on
Supported and Unsupported Activated Hydrotalcites for CO 2 Sorption. Industrial & Engineering
Chemistry Research, 49(17), 8086-8093.
Miguel, C. V., Trujillano, R., Rives, V., Vicente, M. A., Ferreira, A. F. P., Rodrigues, A. E., . . . Madeira,
L. M. (2014). High temperature CO2 sorption with gallium-substituted and promoted
hydrotalcites. Separation and Purification Technology, 127, 202-211.
Oliveira, E. L. G., Grande, C. A., & Rodrigues, A. E. (2008). CO2 sorption on hydrotalcite and alkali-
modified (K and Cs) hydrotalcites at high temperatures. Separation and Purification
Technology, 62(1), 137-147.
Pergher, S. B. C., Corma, A., & Fornes, V. (1999). Materiales laminares pilareados: preparación y
propiedades. Química Nova, 22, 693-709.
Rahul, R., Satyarthi, J. K., & Srinivas, D. (2011). Lanthanum and zinc incorporated hydrotalcites as
solid base catalysts for biodiesel and biolubricants production. Indian Journal of Chemistry-Part
A InorganicPhysical Theoretical and Analytical, 50(8), 1017.
Reichle, W. T. (1986). Synthesis of anionic clay minerals (mixed metal hydroxides, hydrotalcite). Solid
State Ionics, 22(1), 135-141.
Roelofs, J. C., van Bokhoven, J. A., Van Dillen, A. J., Geus, J. W., & de Jong, K. P. (2002). The thermal
decomposition of Mg–Al hydrotalcites: effects of interlayer anions and characteristics of the
final structure. Chemistry–A European Journal, 8(24), 5571-5579.
Sharma, U., Tyagi, B., & Jasra, R. V. (2008). Synthesis and Characterization of Mg−Al−CO 3 Layered
Double Hydroxide for CO2 Adsorption. Industrial & Engineering Chemistry Research, 47(23),
9588-9595.
Shumaker, J. L., Crofcheck, C., Tackett, S. A., Santillan-Jimenez, E., & Crocker, M. (2007). Biodiesel
production from soybean oil using calcined Li–Al layered double hydroxide catalysts. Catalysis
Letters, 115(1), 56-61.
Tang, L., Yamaguchi, D., Wong, L., Burke, N., & Chiang, K. (2011). The promoting effect of ceria on
Li/MgO catalysts for the oxidative coupling of methane. Catalysis Today, 178(1), 172-180.
Velu, S., Sabde, D. P., Shah, N., & Sivasanker, S. (1998). New Hydrotalcite-like Anionic Clays
Containing Zr4+ in the Layers: Synthesis and Physicochemical Properties. Chemistry of
Materials, 10(11), 3451-3458.
Walspurger, S., Boels, L., Cobden, P. D., Elzinga, G. D., Haije, W. G., & van den Brink, R. W. (2008).
The Crucial Role of the K+–Aluminium Oxide Interaction in K+-Promoted Alumina- and
Hydrotalcite-Based Materials for CO2 Sorption at High Temperatures. ChemSusChem, 1(7),
643-650.
Wang, Q., Tay, H. H., Ng, D. J. W., Chen, L., Liu, Y., Chang, J., . . . Borgna, A. (2010). The Effect of
Trivalent Cations on the Performance of Mg-M-CO3 Layered Double Hydroxides for High-
Temperature CO2 Capture. ChemSusChem, 3(8), 965-973.
Wang, Q., Tay, H. H., Zhong, Z., Luo, J., & Borgna, A. (2012). Synthesis of high-temperature CO2
adsorbents from organo-layered double hydroxides with markedly improved CO 2 capture
capacity. Energy & Environmental Science, 5(6), 7526-7530.
Yavuz, C. T., Shinall, B. D., Iretskii, A. V., White, M. G., Golden, T., Atilhan, M., . . . Stucky, G. D. (2009).
Markedly Improved CO2 Capture Efficiency and Stability of Gallium Substituted Hydrotalcites
at Elevated Temperatures. Chemistry of Materials, 21(15), 3473-3475.
Zhang, J., Zhao, N., Wei, W., & Sun, Y. (2010). Partial oxidation of methane over Ni/Mg/Al/La mixed
oxides prepared from layered double hydrotalcites. International Journal of Hydrogen Energy,
35(21), 11776-11786.
201
Life Cycle Assessment of biogas desulfurization by

different technologies to valorize the Biogas
potential energy in sewage plants
Patricio I. Cano1*, Joan Colón2, Martín Ramírez1, Javier Lafuente3, David Gabriel3 and
Domingo Cantero1
1 Department of Chemical Engineering and Food Technologies. Faculty of Sciences. University of

Cadiz. 11510 Puerto Real (Cádiz), Spain. Campus de Excelencia Agroalimentario CeiA3;
patricio.cano@uca.es, martin.ramirez@uca.es, domingo.cantero@uca.es.
2 BETA Technology Centre: “U Science Tech”, University of Vic-Central University of Catalonia,
08500 Vic, Barcelona, Spain; joan.colon@uvic.cat.
3 Department of Chemical, Biological and Environmental Engineering, School of Engineering,
Universitat Autònoma de Barcelona, 08193 Bellaterra, Barcelona, Spain; Javier.Lafuente@uab.cat,
david.gabriel@uab.cat.
ABSTRACT
The goal of this study is to compare, by means of a Life Cycle Assessment (LCA), the
environmental impact of four common desulfurization technologies in sewage treatment plants
(STP) for the removal of hydrogen sulfide (H2S) in biogas considering their operation,
emissions, infrastructures and infrastructures disposal. Furthermore, an extra scenario (no
desulfurization scenario with biogas combustion) was compared with biogas desulfurization
plus energy valorization. Concerning to the four scenarios considered, two biological
processes (aerobic and anoxic biotrickling filter (BTF) and two conventional physical-chemical
processes (a caustic chemical scrubber and impregnated activated carbon) were studied.
Overall, results point out that biological technologies are much more favorable in most
environmental impact categories than conventional technologies and, of course, than the non-
treatment scenario. Concerning to climate change category (carbon dioxide (CO2) equivalent
emission), aerobic desulfurization would emit 5.24 kg CO2 per hour while the anoxic
desulfurization would emit 11.40 kg CO2 per hour. The latter could be improved to emit only
2.71 kg CO2 per hour if nitrate (NO3-) is already present in the sewage treatment plant.
Comparatively, caustic chemical scrubber desulfurization would emit 10.67 kg CO2 per hour
and impregnated activated carbon desulfurization would emit 15.75 kg CO2 per hour.
Moreover, it was detected that for long-term (20 years), the required infrastructures and their
disposal have poor importance in the LCA for all scenarios.
Keywords: biogas; life cycle assessment; desulfurization; energy recovery.
1. INTRODUCTION
Nowadays Global Warming (GW) is becoming a more a threating matter that deteriorates
human life and environment. GW is induced by greenhouse gases such methane (CH4) and
CO2. Both gases have an inducing power on GW that is proved and recognized by all
institutions that fight against GW. CH4 has much more impact on GW than CO2 (25 times
more), so an emission of CO2 is desired instead of CH4 (IPCC 2007). One of the CH4 and CO2
source is biogas which could be produced in many places as in a STP (Deublein and
Steinhauser, 2008). The high contain of CH4 makes biogas interesting to valorize because of
its calorific power. As well, this biogas has H2S and ammonia (NH3), two compounds that
should be removed in order to satisfy European Directives and avoid other environmental
troubles (Directive 2010/75/EU). The biogas purification can be performed by physical-
chemical and biological procedures. Physical-chemical procedures have been carried out
202
traditionally, but now biological procedures are becoming increasingly important because
biological procedures are more economical and environmental friendly than physical-chemical
procedures (IEA Bioenergy, 2009; Tomas et al., 2009). The knowledge about how a procedure
affects on the environment can be measured by a Life Cycle Assessment (LCA). There are
many guidelines to instruct how to develop a LCA, but all of them are based on the same basic
norms (Consoli et al., 1993). LCAs are standardized following UNE-ISO 14040 and 14044. All
LCAs have to respect these standards to be considered as a real LCA and to provide
information that could be useful to improve procedures (International Organization for
Standardization, 2006).
In this study, a comparative LCA study among four different desulfurization technologies
(aerobic biotrickling filter (BTF), anoxic BTF, chemical scrubber with hydrogen peroxide and
activated carbon adsorption) were carried out to select the most suitable technology from a
sustainable point of view. Moreover, these desulfurization technologies were compared with
the “not treatment of H2S without energy recovery”.

2.1 System description
It has been selected a small STP which has an anaerobic digester to remove organic material
from its sludge. The biogas flow rate was of 100 m 3 h-1 and the biogas composition was (%
v/v): CH4 (61.93%), H2S (0.3%), CO2 (37.50%), N2 (0.25%) and NH3 (2.50·10-2). So, the biogas
can not be valorized because of the high H2S presence. Four different scenarios are boarded
to study environmental impacts of the desulfurization processes. The Figure 1 shows the
process flow of selected scenarios, all of them are feed with biogas from the anaerobic
digester. The aerobic BTF, anoxic BTF, caustic chemical scrubber and impregnated activated
carbon were the scenarios 1, 2, 3 and 4, respectively.
Electricity Treated water
Sewage water
grid
Air -
Liquid
Liquid purge emissions -
Sludge
SCENARIO 1
NaOH Air
SCENARIO 2
Raw material
NaHCO3
Biogas
(CH4, CO2,
H2S, NH3
Air
Energy input Energy output
SCENARIO 3 Raw material
H2O2
NaHCO3 NaOH
Activated carbon
SCENARIO 4
KOH
S0, iom ss,
Solid emissions
cti ted om ustion
c r on engine
ANAEROBIC
DESULPHURIZATION COMBUSTION
DIGESTION
Figure 1. Different scenarios that have been boarded to desulfurize the biogas and in this way to
valorize its calorific power
203
The solid emissions related to the operation of all scenarios were dumped in landfill. The liquid
purge of scenarios 1, 2 and 3 was drained to the STP. Because of the sulfur oxidation, sulfate
(SO4-2) would be emitted with this liquid. Finally, the biogas without H2S would be leaded to a
combustion engine, producing electricity that could be used for the STP requirements.
Furthermore, the heat that produces the plant would be considered as a waste because up to
now the trending has been to produce only electricity (Eurobserver, 2014). Finally, some gases
such as SO2 and CO2 would be produced in the combustion engine and emitted to the
atmosphere.
2.1.1 Aerobic BTF
The scenario 1 is the Aerobic BTF (Fig. 1). Aerobic biodesulfurization with BTF has been
proven as an effective technique to remove H2S at pilot scale (Montebello et al., 2013).
The biogas stream would be coupled to an aerobic BTF (total volume of 10 m3 and packed
bed volume of 5 m3) constructed with glass fiber reinforced plastic (GFRP). Furthermore, it
would be required auxiliary equipment such compressor, pumps and pipeline and energy to
operate. Air would be taken from the atmosphere as oxygen (O2) supply for the aerobic BTF
and to operate.
The liquid that would leave the aerobic BTF would be slightly acid (pH 5.5). Because of the
sulfur oxidation, elemental sulfur (S0) would be accumulated and sent to a landfill with the
biomass. It is supposed that a part of the S0 (40%) would be emitted to the atmosphere as
H2S and the rest returns to the biogas stream as H2S. It is supposed that the NH3 in the biogas
stream would be transferred from the gas phase to the liquid phase, emitting NH4+ in the liquid
emissions.
2.1.2 Anoxic BTF
The scenario 2 is the Aerobic BTF (Fig. 1). Anoxic biodesulfurization with BTF has been
proven as an effective technique to remove H2S at pilot scale (Almenglo et al., 2016).
The biogas stream would be coupled to an aerobic BTF (total volume of 10 m3 and packed
bed volume of 5 m3) constructed with GFRP. The NO3- that is required for the desulfurization
process would be supplied by a nitrification reactor of 8.3 m3 that must be constructed with
GRFP to produce nitrate from the NH4+ that is present in the STP. For instance, NH4+ could
be taken from the reject water.
Furthermore, it would be required auxiliary equipment such compressor, pumps and pipeline
and energy to operate. Additionally, air would be taken from the atmosphere as oxygen supply
for the nitrification reactor and for the combustion engine to operate. Besides sodium
hydroxide (NaOH) and sodium hydrogen carbonate (NaHCO3) would be supplied in the
nitrification reactor.
Because of the sulfur oxidation, elemental sulfur (S0) would be accumulated and sent to a
landfill with the biomass. It is supposed that a part of the S0 (40%) would be emitted to the
atmosphere as H2S and the rest of the S0 would be recuperated as biogas. It is supposed that
the NH3 in the biogas stream would be transferred from the gas phase to the liquid phase,
emitting NH4+ in the liquid emissions.
204
2.1.3 Chemical scrubber to desulfurize biogas
The scenario 3 is the desulfurization using a chemical scrubber (Fig. 1). Desulfurization by
chemical scrubber using hydrogen peroxide (H2O2) has been proven an effective technique to
remove H2S at pilot scale (Krischan et al., 2012).
The biogas stream would be coupled to a chemical scrubber BTF (total volume of 1.69 m3 and
packed bed m3of 0.84 m3) constructed with GRFP. Furthermore, it would be required auxiliary
equipment such compressor, pumps and pipeline and energy to operate. Besides, H2O2,
NaHCO3 and NaOH would be supplied to oxidize the H2S, to avoid CO2 transfer from the gas
phase to liquid phase and to neutralize the pH in the liquid phase.
The solubility of NH3 in water is quite high, so it was supposed that all the NH3 was transferred
from the gas phase to the liquid phase, emitting NH4+ in the liquid emissions.
2.1.4 Activated carbon to desulfurize biogas
The scenario 4 is the desulfurization using activated carbon (Fig. 1). Desulfurization by
activated carbon has been proven an effective technique to remove H2S at industrial scale.
The biogas stream would be coupled to an activated carbon desulfurization unit with a packed
be volume of 5.99 m3 constructed with GRFP. Besides, the activated carbon should be
impregnated with potassium hydroxide (KOH) at high temperature (800ºC) with 5 wt% KOH
to activate it (Wu et al., 2005; Kawano et al., 2008). Furthermore, it would be required auxiliary
equipment such compressors and pipeline and energy to operate.
H2S and NH3 would be adsorbed on the activated carbon. Activated carbon would be sent to
a landfill, it is supposed that a part of the H2S and NH3 on the activated carbon (40%) would
be emitted to the atmosphere and the rest of the H2S and NH3 would be returns to biogas
stream.
2.1.5 Not desulfurization
The scenario 5 is the not desulfurization. In the case that to implement a desulfurization
procedure is not economical and environmentally viable, the current situation (not treatment
of the biogas) should continue. This facility would possess the previous anaerobic digester.
The biogas stream would be leaded to a flare of steel to combust the CH4 contained in the
biogas stream. The flare would need air in excess taken from the atmosphere to combust
almost 100% of the CH4 contained in the biogas stream. In the flare it could not use the power
that generates, thus it would be produced heat waste and some gases such as SO2, nitrogen
oxides (NOx) and CO2 and emitted to the atmosphere.
2.2 Life cycle assessment
2.2.1 Goal and Scope
2.2.1.1 Goal of the study
The main goal of this study was to assess which scenario is more optimum for the H2S
removal. Then, the border of each scenario is extended with it the combustion engine to verify
205
if it is necessary or no to set up a desulfurization unit to valorize biogas potential energy when

there is a STP which possesses an anaerobic digester producing biogas.
2.2.1.2 Functional unit
The Functional Unit (FU) is a key parameter to perfume the LCA and it is used to refer all the
inputs and outputs in the same units. In this study, the FU used was 100 m h-1 of biogas at
inlet concentration of 3,000 ppmv of H2S.
2.2.1.3 System boundaries
The goal is to compare the different technologies to desulfurize the biogas. Thereby, the limits
of the four different scenarios are fixed for what is necessary to desulfurize the biogas. To
implement the desulfurization procedure is necessary to build a reactor where the H2S is
removed. The reactor requires auxiliary equipment to work, thus the infrastructures are the
main reactor with its auxiliary equipment. After building the infrastructure, chemicals, energy
and natural sources should be supplied to operate, so the chemicals, energy and natural
sources that are supplied should be included as well within the limits. Additionally, the anoxic
BTF requires NO3- to operate and this NO3- is supplied by the nitrification reactor, thereby, the
infrastructures of the nitrification reactor and its operation should be included in the anoxic
limits as well.
Figure 1 showed the limits of the different scenarios. For the further studies, combustion
engine must be included too, thus combustion engine limits are in the dark zone but separated
with discontinuous points.
2.2.1.4 Life cycle inventory assumptions
It is assumed that the potential damages in the environment due to a desulfurization procedure
are induced because of the construction of the desulfurization unit and its auxiliary equipment;
its usage of chemicals and energy during its daily operation and its emissions to soil, liquid
and air. Transport of solids emissions to the landfill have not been considered because it is
assumed nearness between the STP and the landfill.
It has been assumed that all plants were designed to last a long time operating. Thus, time of
life is equal to 20 years.
2.2.1.4.1 Sulfur transformations, nitrate production and carbon activation
The reactions involved in each scenario are described below.
The aerobic biodesulfurization kinetic (scenario 1) is described by Kuenen et al. (1975).

According to this kinetic, two stages are involved in H2S removal: the first stage is the S0
production (equation 2) and the second stage is the SO4-2 production (equation 3).
H 2 S H  HS  (1)
 
HS 0.5O2 S OH
0
(2)
 2 
HS 2O2  SO 4 H (3)
For the inlet boarded in this study, it has been proven that is possible to remove 99.9% of the
H2S with a SO4-2 production of 80%, while the S0 production is the 20% remaining (López et
206
al., 2016). The anoxic biodesulfurization kinetics (scenario 2) is quite similar to aerobic one
because the two stages that were previously mentioned happen as well and is described by
Mora et al. (2015). Two stages happen, the first one is the S0 production (equation 4) and the
second one is the SO4-2 production from the S0 (equation 5). As in the aerobic
biodesulfurization, the S0 and SO4-2 production depends on the NO3- that is supplied.
H 2 S H  HS  (1)
HS  0.350 NO3  1.40H   0.059 HCO3  0.004CO2  0.013NH 4 
0.013C5 H 7 O2 N  1.21H 2 O  0.175 N 2  S 0 (4)
  
S  0.876 NO  0.343H 2 O  0.379HCO  0.023CO2  0.080 NH 
0
3 3 4
0.080C5 H 7 O2 N  0.824.H   0.44 N 2  SO42 (5)
For the inlet boarded in this study, it has been proven that is possible to remove 99.9% of the
H2S with a SO4-2 production of 69%, while the S0 production is the 31% remaining (Cano et
al., 2015). The anoxic biodesulfurization is coupled to a nitrification reactor, the nitrification
reactor transforms NH4+ into NO3- (Jubany et al., 2008). Two stages happen in NH4+ oxidation,
the first stage is the nitrite (NO2-) generation from NH4+ (equation 6) and the second stage is
the NO3- generation from NO2- (equation 7).
NH 4  1.5O2  NO2  2H   H 2 O (6)

 
NO  0.5O2  NO
2 3 (7)
It is supposed that NH4+ is completely transformed into NO3- (Jubany et al., 2008). The H2S
removal by an oxidation with H2O2 (scenario 3) is described by Couvert et al. (2008). According
to this kinetic, three stages are involved in H2S removal: the first stage is the S0 production
(equation 8), the second stage is the sulfite (SO3-2) production (equation 9) and the final stage
is the SO4-2 production (equation 10). For the inlet boarded in this study, it has been proven
that is possible to remove 99.9% of the H2S transforming all the H2S into SO4-2 because of
operating with excess of H2S (Krischan et al., 2012).
HS   H 2 O2 S 0  H 2 O  OH  (8)
2 
S  H 2 O2  SO
0
3  H 2 O  2H (9)
SO32  H 2 O2  SO42  H 2 O (10)
2 2
SO 3  S  S 2O
0
3 (11)
The H2S removal in the scenario 4 was carried out by the activated carbon. In order to increase
the capacity of the carbon to capture H2S is required to activate with KOH as showed below.
2 KOH  K 2 O  H 2 O (at high temperature, 800ºC) (12)

K 2 O  C  CO  2K (13)
K 2 O  CO 2  K 2 CO3 (14)
K 2 CO3  2C  2 K  3CO (15)
After happening the previous reactions, a porous structure has been developed in the carbon.
This carbon is able to remove H2S and NH3.
207
2.2.1.4.2 STP assumptions
It must not be assumed future damages in the environment due to pollutants present in
wastewater because these damages are attributed to other procedures, out of the current
limits. Thus, it is considered that the possible impacts produced for the liquid is resulting by
the liquid emissions produced by each technology.
2.2.1.4.3 System expansion for electricity production
At first, LCA of different desulfurization technologies were performed to compare different

technologies for biogas desulfurization. But at the same time, a new LCA could be performed
to compare the previous situation (biogas not treatment with combustion in a flare) with biogas
desulfurization to valorize its potential energy.
2.2.1.4 Data collection and data quality
The data origin could be divided into different sections: data obtained directly from
experimental work of the groups involve in the study, data from literature and Institutions
publishing and data provided by companies.
2.2.1.6 Allocation
All the scenarios in this study could be considered as complex systems. In these scenarios is
quite difficult to define the burden allocation procedures. To solve this problem in the present
study, it has been used the “cut-off” methodology defined by Ekvall and Tillman (1997).
2.2.2 Life cycle inventory
In Table 1 the inventory to carry out the LCA is showed, the amount that is explicitly required
for the desulfurization procedures is presented in brackets.
2.2.3 Life cycle impact assessment
The different LCA were run with SimaPro 8.0.4.30 using inventory from the non-profitable
company Ecoinvent 3 (Swiss Center for Life Cycle Inventories, 2013). The LCA was driven
using Recipe Midpoint (E) (Guinée et al., 2002).

3.1 Environmental assessment of the desulfurization technologies
The different LCAs for the desulfurization processes were performed. The result obtained in
each category per FU were the sum of “the construction of the desulfurization unit” (CDU) plus
“the daily desulfurization operations” (DDO) and plus “the final disposal” (FDD). In Figure 2 it
is represented for each scenario the contribution of each stage for the most important
categories.
208
Table 1. Life cycle inventory

Chemical Activated
Units (UF-1) Aerobic Anoxic Not treatment
scrubber carbon
Inputs
Material
Construction
Steel g 120.33 (0.58) 120.33 (0.58) 120.33 (0.58) 120.33 (0.58) 32.08
GRFP g 3.24 (3.24) 4.84 (4.84) 0.12 (0.12) 0.65 (0.65) -
PVC g 0.04 (0.04) 0.04 (0.04) 0.04 (0.04) 0.04 (0.04) -
Polypropylene g 0.46 (0.46) 0.46 (0.46) 0.23 (0.23) 0.23 (0.23) -
Packing material
Polypropylene g 1.48 (1.48) 1.48 (1.48) 0.25 (0.25) - -
Activated carbon
Non-regenerative kg - - - 1.61 (1.61) -
Daily operations
Chemicals
NaOH moles - 1.18 (1.18) 73.57 (73.57) - -
6.93·10-3 (6.93· 5,273.19
NaHCO3 moles - - -
10-3) (5,273.19)
H2O2 kg - - 3.88 (3.88) - -
KOH g - - - 80.37 (80.37) -
Water
Sewage water m3 4.8 (4.8) 0.24 (0.24) 4.39 (4.39) - -
Energy
Electricity kWh 104.00 (8.00) 106.71 (10.71) 100.00 (4.00) - 96.00
Heat kWh - - - 0.32 (0.32) -
Natural Sources
Air m3 688.81 (10.40) 707.29 (29.2) 678.23 (-) 678.23 (-) 678.23
Outputs
Daily operations
Gaseous emissions
167,046.02
H2S mg 37.43 (33.26) 54.58 (50.41) 125.37 (-) 4,179.07
(166,995.55)
CH4 kg 0.41 (-) 0.41 (-) 0.41 (-) 0.41 (-) 0.41 (-)
NH3 g - - - 6.96 (6,96) 0.17
NOx g - - - 0.64 (-) 46.54
SO2 g 0.78 (-) 0.78 (-) 23.33 (-) 11.42 (-) 777.73
Liquid emissions
O2 g 38.40 (-) - 35.11 (-) - -
942.33 814.30 1,142.57
SO4-2 g - -
(942.33) (814.30) (1,142.57)
NH4+ g 18.42 (18.42) 18.42 (18.42) 18.42 (18.42) - -
NO3- mg - 6.70 (6.70) - - -
H+ mg 15.29 (15.29) - - - -
Energy
Electricity kWh 174.45 (-) 174.45 (-) 174.45 (-) 174.45 (-) -
Energy waste
Heat waste kWh 261.68 (-) 261.68 (-) 261.68 (-) 261.68 (-) 407.05
Waste dumped
Biomass g 20.69 (20.69) 31.35 (31.35) - - -
Steel g 120.33 (0.58) 120.33 (0.58) 120.33 (0.58) 120.33 (0.58) 32.08
GRFP g 3.24 (3.24) 4.84 (4.84) 0.12 (0.12) 0.65 (0.65) -
PVC g 0.04 (0.04) 0.04 (0.04) 0.04 (0.04) 0.04 (0.04) -
Polypropylene g 1.94 (1.94) 1.94 (1.94) 0.48 (0.48) 0.23 (0.23) -
In Figure 2 it is possible to see that the most important stage for the different desulfurization
scenarios was DDO. Moreover, for the two conventional physical-chemical processes, the
contribution attributable to CDU and to FDD was almost inconsiderable. However, for the two
biological scenarios, the contribution attributable to CDU and to FDD was measurable but it
was not as considerable as for DDO. For scenario 1 and 2, there were several impact
categories generated mainly because of DDO. As well there was an impact category
generated due to the CDU (“Agricultural Land occupation Category”). Other environmental
effects were generated because of the combination of DDO, principally, plus CDU
(“Freshwater eutrophication Category” and “Metal depletion Category”) or because of the
combination of CDU plus FDD (“Urban land occupation Category”). There were also
environmental effects generated because of the combination of the three stages (“Freshwater
ecotoxicity Category” and “Marine ecotoxicity Category”). Finally, the Natural land
transformation was due to the CDU but it was slightly neutralized by FDD because, in both
scenarios, FDD had a positive effect on the environment. As well for both scenarios, the CDU
had a positive effect on water depletion but it was almost insignificant, in regard to DDO.
209
Figure 2. Each stage contribution on the main impact categories. Black: CDU; Red: DDO; Green: FDD.
For scenario 3 and 4, most impact categories were generated due to DDO. For scenario 3, all
impact categories were mainly generated because of DDO. For scenario 4, all impact
categories were mainly generated because of DDO, but it was also possible to determine
slight influences due to CDU (“Agricultural land occupation Category” and “Metal depletion
Category”).
3.2 Environmental assessment of the possibilities that can appears for the anoxic
desulfurization
For the anoxic desulfurization, it was considered that would be necessary to build an extra
reactor (the nitrification reactor). This reactor would produce nitrate from the NH 4+ that would
be present in the STP. Because of that, the nitrification process, the construction of the
nitrification reactor and its disposal were considered in the LCA. However, other alternatives
could be considered. The alternatives that can be more potential are: A nitrification reactor is
present in the STP, so a NO3- stream is available to be used (scenario 2.1) or using commercial
NO3- (scenario 2.2). Scenario 2.1 is based on implementing anoxic desulfurization in a STP
where a nitrification reactor was before operating. Thus nitrification operation, the nitrification
reactor construction and its disposal would be not considered, minimizing the CO2 up to 2.71
kg CO2 equivalent per FU. Therefore, there would be a decrease of 76.27% per FU for the
impact category “Climate Change Category”. Not only, “Climate Change Category” would be
210
improved, other impact categories such as “Terrestrial Acidification Category”, “Ozone

Depletion Category”, “Human Toxicity Category”, “Ionising radiation Category” or “Water
Depletion Category” would be decreased up to values around 76% per FU. Therefore, in this
case, anoxic desulfurization could be a good alternative to aerobic desulfurization.
Table 2. The results of the different desulfurization technologies

Chemical Activated
Impact category Units (UF-1) Aerobic Anoxic
scrubber carbon
Climate Change 5.240 11.395 10.666 15.745
Construction of the Desulfurization Unit 0.037 0.052 0.004 0.008
kg CO2
Daily Desulfurization Operations 5.202 11.343 10.663 15.737
Final Disposal of the Desulfurization Process 0.000 0.000 0.000 0.000
Ozone Depletion 0.955 2.085 0.951 0.385
Construction of the Desulfurization Unit mg CFC-11 0.001 0.002 0.000 0.000
Daily Desulfurization Operations eq 0.954 2.084 0.951 0.385
Terrestrial acidification 38.235 83.361 64.527 40.499
g SO2 eq
Freshwater eutrophication 55.924 86.111 1,442.701 81.751
mg P eq
Daily Desulfurization Operations 49.285 77.499 1,441.440 80.441
Marine eutrophication 15.027 17.229 22.442 1.410
g N eq
Human toxicity 6.951 14.891 78.793 5.722
Construction of the Desulfurization Unit kg 1-4-DB 0.224 0.297 0.064 0.042
Terrestrial ecotoxicity 731.741 1,606.322 4,176.673 413.576
Construction of the Desulfurization Unit mg 1,4-DB 7.383 10.655 3.225 2.011
Daily Desulfurization Operations eq 723.770 1,594.835 4,173.408 411.445
Freshwater ecotoxicity 1.891 3.693 129.466 3.521
g 1,4-DB eq
Marine ecotoxicity 2.907 5.909 133.091 3.576
Construction of the Desulfurization Unit kg 1,4-DB 0.212 0.280 0.065 0.043
Water depletion 12.704 28.339 266.931 19.587
Construction of the Desulfurization Unit -0.296 -0.132 0.015 0.094
l
Metal depletion 3.030 5.382 371.975 13.764
g Fe eq
Fossil depletion 1,307.550 2,842.258 3,307.314 3,996.444
g oil eq
Daily Desulfurization Operations 1,293.125 2,822.884 3,305.720 3,993.676
Scenario 2.2 is based on using commercial NO3-. Thus nitrification operation, the nitrification
reactor construction and its disposal would be not considered, minimizing the CO2 up to 8.23
kg CO2 equivalent per FU. Therefore, there would be a decrease of 27.82% per FU for the
impact category “Climate Change Category”. Not only, “Climate Change Category” would be
improved, other impact categories such as “Terrestrial Acidification” or “Ozone Depletion”
would be decreased, up to 57.88 per FU for “Ozone Depletion Category” and 36.13% per FU
for “Terrestrial Acidification Category”. Despite of these results, other categories would be
deteriorated too much. For example, “Freshwater Eutrophication Category” would increase up
to 1,180.92% per FU; “Human Toxicity” up to 351.95% per FU and “Metal Depletion Category”
up to 7,210.60% per FU. Therefore, in this case this alternative could not be a realistic
alternative to aerobic desulfurization.
211
3.3 Environmental assessment of biogas combustion
The different LCAs for the biogas combustion were performed. There were 5 different
scenarios. 4 scenarios were desulfurization process plus combustion, valorizing the electricity
produced. The fifth scenario was the combustion of the biogas without energy recovery. For
all cases with biogas desulfurized, most of the impacts on the environment were positive. The
negative impacts on the environment were “Freshwater Eutrophication Category”, “Marine
Eutrophication Category”, “Freshwater Ecotoxicity Category”, “Marine Ecotoxicity Category”,
“Agricultural land occupation Category”, “Urban land occupation Category”, “Natural land
transformation Category” and “Metal depletion Category”. Additionally only for Scenario 3,
“Human toxicity Category” and “Water depletion Category” would be negative because the
negative part, associated to the desulfurization process, would be bigger than the positive
part, associated to the energy recovery part. On the other part, for scenario 5, all categories
would be negative because of the no energy recovery due to the SOx presence. In Figure 3 it
is represented for each scenario the contribution of each stage for the scenario 1 and for
scenario 5. There are 6 stages, the 3 already present for the desulfurization cases plus “the
construction of the combustion unit” (CCU) plus “the daily combustion operations” (DCO) and
plus “the final disposal after combusting” (FDC). When a stage it is under 0, the meaning is
this stage is positive for the environment. It is clearly checked that the combustion stage for
energy recovery makes most impact categories positives for the environment. For scenario 2,
3 and 4, the final results of integrating the combustion unit with the desulfurization unit are
also positives but the values are lower because of the major desulfurization impacts.
However, the combustion without energy recovery is very harmful for the environment as
showed in Figure 3. For “Climate Change Category”, the emissions of CO2 equivalent per FU
would be of 65.24 kg, while with a previous desulfurization it would be possible to save CO 2,
up to 36.93 kg CO2 per FU for scenario 1, 33.32 kg CO2 per FU for scenario 2, 34.01 kg CO2
per FU for scenario 3 and 30.13 kg CO2 per FU for scenario 4. It is also important to indicate
the results obtained in Terrestrial acidification Category. It would be possible to save up to
297.12 g SO2 per FU for scenario 1, 271.10 g SO2 per FU for scenario 2, 266.78 g SO2 per FU
for scenario 3 and 302.91 g SO2 per FU for scenario 4. On the other hand, operating without
energy recovery, it would suppose emissions of 1,268.19 g SO2 per FU.
Figure 3. Each stage contribution on the impact categories for “Aerobic Desulfurization plus Biogas
valorization” and “Not biogas desulfurization plus Combustion”. Black: CDU; Red: DDO; Green: FDD;
Yellow: CCU; Blue: DCO; Pink: FDC.
212
4. CONCLUSIONS
The performance of the LCA for 4 different biogas desulfurization processes in STP has
determined that aerobic desulfurization was the most appropriate process. However, anoxic
desulfurization could be a convincing alternative if NO3- would be already present in the STP.
As well, desulfurization by chemical scrubber have several impacts on the environment that
should be corrected to be environmental-friendly. Finally, desulfurization by activated carbon
is inadvisable because of the high amounts of CO2 emitted, compared with the rest of
desulfurization technologies. It is proven that biogas desulfurization is necessary for an
environmental improvement and biogas combustion without energy recovery is a process that
is not recommendable because it is very harmful for the environment.
ACKNOWLEDGMENTs
The Spanish Government has provided financial support to carry out the research work
through the project CTM2012-37927-C03/FEDER “Monitoring, modelling and control towards
the optimization of anoxic and aerobic desulfurizing biotrickling filters”.
Department of Chemical, Biological and Environmental Engineering (Universitat Autònoma de

Barcelona) is a unit of Biochemical Engineering of the Xarxa de Referència en Biotecnologia
de Catalunya (XRB), Generalitat de Catalunya.
REFERENCES
Almenglo F, Ramírez M, Gómez J M, Cantero D (2016). Operational conditions for start-up

and nitrate-feeding in an anoxic biotrickling filtration process at pilot scale. Chemical
Engineering Journal 285, 83–91.
Couvert A, Sanchez C, Laplanche A, Renner C (2008). Scrubbing intensification for sulphur
and ammonia compounds removal. Chemosphere 70 (8), 1510–1517
Ekvall T, Tillman A (1997). Open-Loop recycling: criteria for allocation procedures.
International Journal of Life Cycle Assessment 2 (3), 155–162.
Jubany I, Carrera J, Lafuente J, Baeza J A (2008). Start-up of a nitrification system with
automatic control to treat highly concentrated ammonium wastewater: Experimental
results and modelling. Chemical Engineering Journal 144, 407-419.
Kawano T, Kubota M, Onyango M, Watanabe F, Matsuda H (2008). Preparation of activated
carbon from petroleum coke by KOH chemical activation for adsorption heat pump.
Applied Thermal Engineering 28 (8-9), 865-871.
Krischan J, Makaruk A, Harasek M (2012). Design and scale-up of an oxidative scrubbing
process for the selective removal of hydrogen sulfide from biogas. Journal of
Hazardous Materials 215-216, 49-56
Kuenen J G (1975). Colourless sulphur bacteria and their role in the sulphur cycle. Plant Soil
43 (1), 49–76.
López L R, Bezerra T, Mora M, Lafuente J, Gabriel D (2016). Influence of trickling liquid
velocity and flow pattern in the improvement of oxygen transport in aerobic biotrickling
filters for biogas desulfurization. Journal of Chemical and Biotechnology 91 (4), 1031-
1039.
Montebello A, Bezerra T, Rovira R, Rago L, Lafuente J, Gamisans X, Campoy S, Baeza M,
Gabriel D (2013)
Operational aspects, pH transition and microbial shifts of a H2S desulfurizing biotrickling filter
with random packing material. Chemosphere 93 (11), 2675-2682
Mora M, Fernández M, Gómez J M, Cantero D, Lafuente J, Gamisans X, Gabriel D (2015).
Kinetic and stoichiometric characterization of anoxic sulfide oxidation by SO-NR mixed
cultures from anoxic biotrickling filters. Applied Microbiology and Biotechnology 99 (1),
77-87.
213
Wu M, Zha Q, Qiu J, Han X, Guo Y, Li Z, Yuan A, Sun X (2005). Preparation of porous carbons
from petroleum coke by different activation methods. Fuel 84 (14-15), 1992–1997
Cano P I, Brito J, Almenglo F, Ramirez M, Gómez J M, Cantero D (2015). Optimization of
nitrate dosage in anoxic biotrickling filters for biogas desulfurization. 7th International
conference on biotechniques for air pollution control. September 2-4. Gent, Belgium,
183–192.
Eurobserver (2014). Biogas Barometer. (Cited 25/4/2016). Available from:
http://www.eurobserv-er.org/pdf/biogas-barometer-2014-en/
Intergovernmental Panel on Climate Change (IPCC) (2014). Synthesis Report. Cambridge
University Press, Cambridge, United Kingdom and New York, NY, USA.
Petersson A, Wellinger A (2009). Biogas Upgrading technologies- Developments and
innovation. IEA Bioenergy, Task 37 (2009). (Cited 25/4/2016). Available from:
http://www.iea-biogas.net/files/daten-redaktion/download/publi-
task37/upgrading_rz_low_final.pdf.
Directive 2010/75/EU of the European Parliament and the Council of 24 November 2010 on
industrial emissions (integrated pollution prevention and control).
ISO 14040:2006 - Environmental management -- Life cycle assessment -- Principles and
framework.
ISO 14044:2006 - Environmental management -- Life cycle assessment -- Requirements and
guidelines.
Consoli, F., Allen, D. Boustead, I., Fava, J., Franklin, W., Jensen, A.A., de Oude, N., Parrish,
R., Perriman, R. Postlethwaite, D., Quay, B., Seguin J. and Vigon, B. (1993) Guidelines
for Life-Cycle Assessment: A ‘Code of Practice”. SETAC. Brussels and Pensacola.
Deublein D, Steinhauser (2008) Biogas from Waste and Renewable Resources: An
Introduction. Wiley-VCH Verlag GmbH & Co. KGaA. Weinheim.
Guinée J, Gorrée M, Heijungs R, Huppes, G, Kleijn R, Koning A, Oers L, Wegener Sleeswijk
A, Suh S, Udo de Haes H A, Bruijn H, Duin R, Huijbregts M (2002) Handbook on Life
Cycle Assessment: Operational Guide to the ISO Standards. Kluwer Academic
Publishers. Dordrecht.
214
Nitrous oxide: global trends, mitigation and

abatement
Osvaldo D. Frutos1,3, Raquel Lebrero1, Guillermo Quijano2, David Marín1, Raúl
Muñoz1*
1Department of Chemical Engineering and Environmental Technology, University of Valladolid, Dr.
Mergelina, s/n, 47011, Valladolid, Spain. Tel. +34 983186424
2Agrarian Engineering School, Soria Campus, University of Valladolid, Soria 42004, Spain. Tel.
3Facultad de Ciencias Agrarias, Universidad Nacional de Asunción, Campus Ciudad de San Lorenzo,
Paraguay. Tel. +595 21585606

*Corresponding author: mutora@iq.uva.es, Tel. +34 983186424.
ABSTRACT
Nitrous oxide (N2O) is the third most important greenhouse gas (GHG), and together with
carbon dioxide and methane, accounts for 98 % of the total GHG emissions worldwide. The
global warming potential of N2O is almost 300 times higher than that of CO2 and constitutes
6.2 % of the total GHG anthropogenic emissions, which represents 3.1 ± 1.9 Gt of CO2-
equivalents per year based on its global warming potential in a 100-year time horizon.
Furthermore, N2O is considered the most critical O3-depleting substance emitted in this XXI
century with an atmospheric concentration increasing trend of 0.3 % per year. Approximately
40 % of the total N2O emissions originate from anthropogenic activities, agriculture, chemical
industries and wastewater treatment (WWT) being its most important sources. Based on the
urgent need to reduce GHG emissions to limit global temperature increase to 2 ºC above pre-
industrial levels, several technologies have been developed to abate N2O from anthropogenic
emissions. In this context, biotechnologies are proposed as promising end of pipe
technologies for the cost-effective and environmentally friendly treatment of this GHG. This
work constitutes a critical assessment of the global problem of N2O emissions and mitigation
strategies to reduce the N2O emitted from WWT processes and nitric acid production plants
based on 2 years of experimentation. The most significant results of two innovative processes
developed at the Department of Chemical Engineering and Environmental Technology of
Valladolid University will be presented.
Keywords: Nitrous oxide, greenhouse gas abatement, biotechnology.
1. INTRODUCTION
The greenhouse gases (GHGs) atmospheric concentration has increased more than 45 %
since the preindustrial era (IPCC, 2014). Nitrous oxide (N2O), the third most important GHG
with a global warming potential 300 times higher than that of CO2 and an atmospheric
persistence of 150 years, accounts for 6.2 % of the total GHGs emitted globally. N 2O is also
one of the main sources of stratospheric NOx and is considered as the most important ozone
depleting substance emitted in this XXI century, with a yearly atmospheric concentration
increase of 0.3 % (Ravishankara et al., 2009). By 2013, the anthropogenic N2O emissions
reached 250.5 Tg of CO2-equivalents (TgCO2-eq) per year in the European Union (EU-28)
despite the continuous reduction efforts. Of this total, 196.8 TgCO2-eq were produced by
agriculture, 30.8 and 11.2 TgCO2-eq were emitted from energy production (combustion and
fugitive emissions) and industrial processes, respectively, and 11.6 TgCO2-eq were generated
in the waste management sector (EEA, 2015). The release of N2O from fugitive sources such
as agriculture activities is difficult to control. Therefore, several studies have been carried out
215
to determine the N2O generation mechanisms and thus propose mitigation strategies (Rees
et al., 2013; Thomson et al., 2012). On the contrary, N2O from industrial and waste
management activities is emitted from stationary sources and the treatment of these emissions
can be carried out by end of pipe technologies when mitigation is not possible. In this sense,
wastewater treatment processes constitute an important source of N2O emission in the IPPC
sector of waste management, with 3.2 % of the total N2O emissions in the EU-28. Moreover,
data from US-Environmental Protection Agency indicate that 2.5 % of the global N2O
emissions in 2005 came from sewage treatment processes (EPA, 2011). On the other hand,
the main source of N2O emissions within the industrial sector are Nitric and adipic acid plants,
which emissions reached up to 500 kTon of N2O per year in spite of the recent abatement
strategies applied in these industries (Kondratenko and Pérez-Ramírez, 2006).
Some physical-chemical technologies, such as non-selective catalytic reduction (NSCR) or

catalytic decomposition, have been applied as end of pipe strategies for the treatment of N2O
emissions from industrial sources. However, these systems require a reducing agent, such as
hydrocarbons or ammonia, and high temperature for N2O destruction in most of the cases.
Conversely, research on N2O emission control from WWTPs has been mainly focused on
minimizing and/or preventing N2O generation and its release to the atmosphere, with limited
application of end of pipe technologies (Law et al., 2012; Richardson et al., 2009; Zheng et
al., 2014; Zhu et al., 2013).
Biological technologies have demonstrated promising characteristics such as robustness, cost

efficiency and environmentally friendliness for the treatment of a wide range of polluted waste
gases (Estrada et al., 2011). However, in spite of these advantages, biotechnological systems
have been scarcely evaluated for the abatement of N2O emissions. In this regards, the sole
biological process for N2O degradation is denitrification, where this GHG is an obligate
intermediate in the reduction steps of NO3- or NO2- to N2 during organic matter oxidation. Thus,
two biological systems, a bioscrubber and a bubble column bioreactor (BCR), have been
evaluated for the continuous abatement of simulated N2O emissions typically found in WWTPs
(Exp. 1) and nitric acid production plants (Exp. 2). This work presents the most relevant results
of two years of research carried out at the Chemical Engineering and Environmental
Technology Department of Valladolid University.
2.1. Chemicals
The synthetic wastewater (SW) used in the Exp. 1 was composed of (in g L-1 of tap water):
peptone 0.16, meat extract 0.11, urea 0.03, NaCl 0.007, CaCl2∙2H2O 0.004, MgSO4∙7H2O
0.002, K2HPO4 0.028, CuCl2∙2H2O 50×10-6 and glucose 0.25. The final total organic carbon
(TOC) and total nitrogen (TN) concentrations in the SS were 256 ± 23 and 54 ± 3 mg L -1,
respectively.
The mineral salt medium (MSM) used in the Exp. 2 was composed of (g L-1): Na2HPO4·12H2O
6.16, KH2PO4 1.52, MgSO4·7H2O 0.2, CaCl2 0.02, NH4Cl 1.5, and 10 mL L−1 of a trace element
solution (containing per liter: EDTA 0.5 g, FeSO4·7H2O 0.2 g, ZnSO4·7H2O, 0.01 g,
MnCl2·4H2O 0.003g, H3BO3 0.03 g, CoCl2·6H2O 0.02 g, CuCl2·2H2O 0.001 g, NiCl2·6H2O
0.002 g, NaMoO4·2H2O, 0.003 g). The final pH of the MSM was 7.
216
2.2. Experimental set-ups and operations
2.2.1 Bioscrubber for WWTPs emissions treatment

The bioscrubber constructed for the experiment of WWTPs N2O abatement consisted of an
absorption column filled with 2 L of Kaldnes rings interconnected to a 3 L fixed bed reactor
(FBR) filled with 1 L of polyurethane foam (PUF) with a Methylotrophic community attached
from a previous experimental set-up (Frutos et al., 2014). The bioscrubber was operated
continuously at a gas empty bed residence time (EBRT) of 3 min for the abatement of a
simulated N2O emission (composed of 100 ± 8.4 ppmv of N2O in air) simultaneously with the
treatment of the SW which acted as the energy and carbon source for N2O reduction by the
microorganisms. The SW was introduced at the bottom of the FBR, where it was mixed with
the N2O-laden recycling liquid from the absorption column, and was further recirculated from
the top of the FBR to the top of the packed bed absorption column using a peristaltic pump.
The SW was supplied to the FBR at flow rates determined by the maintenance of anoxic
conditions in the FBR (dissolved oxygen (DO) concentration = 0 mg L-1). No N2O was supplied
to the inlet air for the first 18 days of operation (stage I) in order to assess any potential N2O
generation in the system as a result of wastewater treatment. During stage I, the bioscrubber
was operated with a SW flow rate of 3 ± 0.1 L d-1 and a liquid recycling velocity (UL) of 1 m h-
1
. Stage II (days 19-51) was characterized by process operation at a N2O of 100 ± 8.4 ppmv,
UL of 1 m h-1 and a SW inlet flow of 4 ± 1 L d-1. UL was increased up to 4 m h-1 during stage III
(days 52- 83) concomitantly with an increase in SW flow rate to 19 ± 1 L d-1. The bioscrubber
was operated from day 84 to 104 (stage IV) with a UL of 8 m h-1 and a SW flow rate of 36 ± 4
L d-1. Similar operational conditions were maintained during stage V (days 105-118), which
was characterized by the supplementation to the SW of CuCl2∙2H2O at 50 µg Cu2+ L-1 in order
to assess the influence of copper (a structural component of the nitrous oxide reductase) on
N2O degradation. Finally, the FBR volume was increased from 3 L to 7.5 L in stage VI (days
119- 135) in order to increase N2O reduction under process operation at a UL of 8 m h-1 and a
SW flow rate of 38 ± 1 L d-1. This change was accomplished transferring the complete content
of the packed bed from the initial FBR to the new one, plus 4 L of new PUF. A summary of the
operational conditions of this experiment is shown in table 1. From day 136 to 140, the EBRT
was stepwise increased to 6, 12, 18, 40 and 80 min under the operating conditions set in stage
VI.
Table 12. Resume of bioscrubber operational conditions

OPERATIONAL CONDITIONS
Addition
SW
N2O inlet of Cu+2 FBR
Stages Operation UL EBRT flow
concentration in the volume
days (m h-1) (min) rate
(ppmv) SW (L)
(L d-1)
(µg L-1)
I 0.05 ± 0.06 0-18 1 3 3 ± 0.1 0 3
II 19-51 1 3 4±1 0 3
III 52-83 4 3 19 ± 1 0 3
IV 100 ± 8.4 84-104 8 3 36 ± 4 0 3
V 105-108 8 3 36 ± 4 50 3
VI 119-135 8 3 38 ± 1 50 7.5
2.2.2 Bubble column for nitric acid plants emissions treatment

The emissions from nitric acid plants are characterized by a typical composition of 100-3500
ppmv of NOx, 300-3500 ppmv of N2O, 1-4 % of O2 and 0.3-2 % of H2O. Hence, a simulated tail
gas composed of N2O at 3377 ± 342 ppmv with 1 ± 0.1 % of oxygen was treated in a BCR of
217
2.3 L inoculated with 0.5 L of Paracoccus denitrificans strain and filled with MSM resulting in
an initial total suspended solid (TSS) concentration of 56 mg L-1. The N2O gas inlet flow rate
was maintained at 110 mL min-1, which corresponded to an EBRT of 17 ± 0.6 min. Pure
methanol (CH3OH) was injected in the gas line by means of a syringe pump in a sample port
filled with fiberglass wool to facilitate its evaporation at a flow rate of 1.9 mL d -1, resulting in a
daily CH3OH loading rate of 331 g m-3 d-1. The system was monitored every two days for
determining CO2, N2O and O2 gas concentration, dissolved CH3OH, TOC, TN and TSS before
the replacement of 300 mL of MSM to replenish nutrients concentration.
2.3. Analytical procedures

A Bruker Scion 436 gas chromatograph (Palo Alto, USA) coupled with an Electron Capture
Detector equipped with a HS-Q packed column (1 m × 2 mm ID × 3,18 mm OD) (Bruker, USA)
was used for N2O gas concentration analysis. The injector, detector and oven temperatures
were maintained at 100 ºC, 300 ºC and 40 ºC, respectively. Nitrogen was used as the carrier
gas at 20 mL min-1.
The CO2 and O2 gas concentration were determined in a Bruker 430 gas chromatograph (Palo
Alto, USA) equipped with a thermal conductivity detector and CP-Molsieve 5A (15 m × 0.53
µm × 15 µm) and a CP-PoraBOND Q (25 m × 0.53 µm × 10 µm) columns. The oven, injector
and detector temperatures were set at 45, 150 and 175 ºC, respectively. Helium was used as
the carrier and make-up gas at 6 and 24 mL min-1, respectively.
A TOC-VCSH analyzer (Shimadzu, Tokyo, Japan) coupled with a total nitrogen

chemiluminescence detection module (TNM-1, Shimadzu, Japan) was used to determine TOC
and TN concentrations. In the FBR, the DO concentration was continuously monitored by a
DO probe connected to a 4100e O2 Transmitter (Metter Toledo, Undorf, Switzerland). The pH
of the FBR was monitored on-line by a pH probe connected to a Black Stone BL 7916 pH
controller (Hanna Instrument, Romania) and the pH of the BCR was periodically monitored
with using a pH/mV/°C meter (pH 510 Eutech Instruments, Nijkerk, the Netherlands). The
determination of TSS concentration was performed according to standard methods (APHA,
2005).
3. RESULTS AND DISCUSSIONS
3.1. Bioscrubber for WWTPs emissions treatment

The pH of the bioscrubber during the 140 days of operation remained constant at 7.3 ± 0.2.
N2O concentrations of 1.3 ± 1 ppmv were recorded at the outlet of the absorption column during
the first 18 days of operation in the absence of N2O supply at a UL of 1 m h-1 (Fig 1A). These
results are in agreement with previous studies reporting N2O production in wastewater
biofiltration (Garzón-Zúñiga et al., 2005; Melse and Mosquera, 2014; Tallec et al., 2006b).
Steady state TOC and TN removals of 90 ± 5 % and 75 ± 15 %, respectively were recorded
in this stage.
Process operation at an inlet N2O gas concentration of 100 ± 8 ppmv and a UL of 1 m h-1 during
stage II was characterized by a steady gas N2O removal efficiency (RE) of 8 ± 3 % (Fig. 1A).
An increase in TOC-RE up to 95 ± 3 % was recorded. Likewise, high TN-RE was observed
(90 ± 2 %) during stage II.
218
The increase in UL to 4 m h-1 during stage III entailed an enhancement in the gas N2O
degradation up to steady state REs of 17 ± 2 % (Fig. 1A). However, a significant deterioration
in the TOC-RE (89 ± 4 %) resulted from the increase of SW flow rate to 19 ± 1 L d -1, which
reduced the hydraulic retention time (HRT) of the SW in the FBR.
An increase in the steady state gas N2O RE up to 26 ± 5 % (Fig. 1A) was recorded with the
increase in UL to 8 m h-1 in stage IV. The increase of the UL in the bioscrubber improved the
overall gas N2O removal as a result of an enhanced N2O transfer and a reduction in the anoxic
tank. On the other hand, the increase in the SW flow rate up to 36 ± 4 L d-1 in order to maintain
anoxic conditions in the FBR mediated a decrease in the TOC-RE to 85 ± 4 %. Nevertheless,
TN removal showed no significant variation under steady state conditions (43 ± 4 %) during
stage IV.
The addition of CuCl2∙2H2O to the SW in stage V showed unexpected results with steady state
gas N2O REs of 29 ± 2 %, which indicates no beneficial effect in the nitrous oxide reductase
production conversely to the findings reported by many authors (Charnock et al., 2000; Farrer
et al., 1991). For instance, Zhu et al. (Zhu et al., 2013) observed a decrease from 50 to 73 %
in the N2O production in an activated sludge process by the addition of 50 to 100 µg L-1 of Cu.
Finally, the increase in the volume of FBR from 3 to 7.5 L during stage VI brought about a gas
N2O RE of 36 ± 3 % (Fig 1A). A TOC removal of 91 ± 1 % was recorded under steady state
operation along with a TN-RE of 42 ± 5 %. The higher FBR size mediated an increase in the
HRT and in the dissolved N2O residence time in the anoxic tank. The increase of these two
parameters promoted an enhancement of the overall system performance.
Figure 41. Time course of (A) bioscrubber performance with inlet (○) and outlet (x) N2O concentration
and removal efficiency (solid line); and (B) evolution of the N2O (●) and TOC (▲) removal with the
stepwise increase in EBRT
Process operation at an EBRT of 3 min supported a gas N2O-RE of 36 ± 0.2 % along with
TOC removals of 92 %. The stepwise increase in the gas EBRT in the absorption column
resulted in a sequential enhancement in the gas N2O-REs but also in the deterioration of TOC
removal (Figure 1B). Hence, the highest gas N2O-RE (93.8 ± 0.2 %) was achieved at an EBRT
of 80 min together with TOC-RE of 74 % (Figure 1B). The deterioration of TOC-RE was an
effect of the reduction in the gas mass transfer rate to the liquid phase due to higher
EBRTs(Dorado et al., 2009; Lebrero et al., 2011), and subsequently, less electron acceptor
was transferred to the FBR for TOC degradation.
219
The bioscrubber studied showed the feasibility of a simultaneous removal of N2O from WWTPs
coupled with the treatment of wastewater. The wastewater was evaluated as a cheap carbon
and electron donor source to support cost-effective N2O removal in an innovative bioprocess
with high level of wastewater depuration. However, despite the promising results, more studies
are needed in order to overcome mass transfer limitations, since high EBRTs (and therefore
high capital cost) were required to reach high N2O REs due the poor solubility of the pollutant.
3.2. Bubble column for nitric acid plants emissions treatment

The pH of the BCR remained at 6.65 ± 0.13 with a DO concentration of 0.11 ± 0.14 mg L -1
during the entire experimentation period. The first ten days of operation were characterized
by a gradual increase in the RE of N2O in the BCR (Figure 2A). This was likely due to the need
for an adaptation period where the P. denitrificans synthetized the enzymes necessary for the
anoxic degradation of CH3OH. The inlet and outlet N2O concentrations of the BCR were 3377
± 342 and 441 ± 74 ppmv, respectively (Figure 2A) with a steady state N2O RE of 87 ± 3 %.The
results here reported represent the highest N2O RE observed during continuous operation (50
days of steady state removal) in biological systems.
The steady state production of CO2 was reached after 10 days of operation, concomitantly
with the stabilization of N2O RE (Figure 2B). The BCR showed a production of CO2 of 308 ±
36 g m-3 d-1. The biomass concentration and TOC concentration increased gradually during
the first 30 days of operation, thereafter steady state TSS concentration of 867 ± 109 mg L-1
(Figure 2B) and TOC concentration 390 ± 38 mg L-1 were recorded until the day 62.
Figure 2. Time course of (A) inlet (x) and outlet (solid line) N2O concentration and N2O RE (◊) in the
BCR and (B) CO2 production (∆) and TSS concentration (x)
220
These results showed that biotechnologies are an interesting alternative for the abatement of
N2O generated in nitric acid production plants compared with their physical-chemical
counterparts. The performance of the BCR regarding RE can be compared with the NSCR
technology which can reach removals of 80-90 % of N2O (Environmental Protection Agency,
2010). However, this technology together with other abatement systems explored nowadays,
such as selective catalytic reduction or catalytic decomposition, entail the production of
secondary emissions (COx) and consumption of reductants (natural gas, ammonia) and
precious metals (Environmental Protection Agency, 2010; Jiang et al., 2009). Alternatively,
the BCR is a biotechnology characterized by the low energy consumption since the agitation
is accomplished pneumatically by the gas phase bubbled from the bottom. Moreover,
simplicity in construction with no moving parts and high gas-liquid mass transfer rates
constitute also key advantages (Fu et al., 2007; Merchuk et al., 1994). The greatest limitation
of the system here proposed was the high gas EBRT (>15 min) required to obtain high
elimination due the poor solubility of N2O, which results in large volume reactors. Therefore,
additional studies devoted to optimize bioreactor design and operational strategies are
necessary in order to overcome the N2O mass transfer limitations typically encountered during
the off-gas abatement of this GHG. Likewise, more studies have to be developed to evaluate
the performance of this biotechnology with higher O2 concentration in the waste gas, since the
N2O reduction by denitrification can be severely affected by the presence of oxygen (Tallec et
al., 2006a).
3. CONCLUSIONS
The studies here presented demonstrated the potential of biotechnologies for the treatment of
N2O emissions from wastewater treatment and nitric acid production plants. In this context,
high N2O removal efficiencies were recorded in the bioscrubber together with high organic
carbon reduction from a model synthetic wastewater. Likewise, the BCR evaluated for the
treatment of emissions from nitric acid production plants showed high N2O removals
comparable to that obtained in physical-chemicals technologies. However, both
biotechnologies presented limitations in the mass transfer of N2O due to its poor solubility.
Therefore, more research focused on the development of innovative systems and operational
strategies are needed to overcome this drawback.
ACKNOWLEDGMENTS
These studies were supported by the Spanish Ministry of Economy and Competitiveness
(CTQ2012-34949 and Red NOVEDAR CTQ2014-51693-REDC projects) and the European
Union through the Erasmus Mundus Program BABEL and FEDER Funding Program.
REFERENCES
APHA, 2005. Standard Methods for the Examination of Water and Wastewater 21st ed.,
Whasington, D.C.: American Public Health Association.
Charnock JM, Dreusch A, Körner H, Neese F, Nelson J, Kannt A, Michel H, Garner CD,
Kroneck P, Zumft W (2000) Structural investigations of the CuA centre of nitrous oxide
reductase from Pseudomonas stutzeri by site-directed mutagenesis and X-ray
absorption spectroscopy. European Journal of Biochemistry 267(5):1368–1381.
Dorado AD, Rodríguez G, Ribera G, Bonsfills A, Gabriel D, Lafuente J, Gamisans X (2009)
Evaluation of mass transfer coefficients in biotrickling filters: Experimental
221
determination and comparison to correlations. Chemical Engineering and Technology

32:1941–1950.
Environmental Protection Agency (EPA) (2010) Available and Emerging Technologies for
Reducing Greenhouse Gas Emissions from the Nitric Acid Production Industry, North
Carolina. Available at: http://www.epa.gov/sites/production/files/2015-
12/documents/nitricacid.pdf.
Environmental Protection Agency (EPA) (2011) Global Anthropogenic Non-CO2 Greenhouse
Gas Emissions: 1990-2030. Available at:
https://www3.epa.gov/climatechange/EPAactivities/economics/nonco2projections.ht
ml
Estrada JM, Kraakman NJ, Muñoz R, Lebrero R (2011) A comparative analysis of odour
treatment technologies in wastewater treatment plants. Environmental science &
technology 45(3):1100–1106.
European Environment Agency (EEA) (2016) EEA greenhouse gas - data viewer. Available
at: http://www.eea.europa.eu/data-and-maps/data/data-viewers/greenhouse-gases-
viewer
Farrer J, Thomson J, Cheesman M, Dooley M, Zumft W (1991) A model of the copper centres
of nitrous oxide reductase (Pseudomonas stutzeri): Evidence from optical, EPR and
MCD spectroscopy. FEBS Letters 294(1-2):11–15.
Frutos OD, Arvelo IA, Pérez R, Quijano G, Muñoz R (2014) Continuous nitrous oxide
abatement in a novel denitrifying off-gas bioscrubber. Applied Microbiology and
Biotechnology 99(8):3695-3706.
Fu C-C, Fan L-S, Wu W-T (2007) Flow Regime Transitions in an Internal-Loop Airlift Reactor.
Chemical Engineering & Technology 30(8):1077–1082.
Garzón-Zúñiga M, Lessard P, Aubry G, Buelna G (2005) Nitrogen elimination mechanisms in
an organic media aerated biofilter treating pig manure. Environmental technology
26(4):361–371.
Intergovernmental Panel on Climate Change (IPCC) (2013) Summary for Policymakers. In:
Climate Change 2013: The Physical Science Basis. Contribution of Working Group I
to the Fifth Assessment Report of the Intergovernmental Panel on Climate Change
[Stocker, T.F., D. Qin, G.-K. Plattner, M. Tignor, S.K. Allen, J. Boschung, A. Nauels, Y.
Xia, V. Bex and P.M. Midgley (eds.)]. Cambridge University Press. pp. 1–30
Jiang H, Wang H, Liang F, Werth S, Schiestel T, Caro J (2009) Direct decomposition of nitrous
oxide to nitrogen by in situ oxygen removal with a perovskite membrane. Angewandte
Chemie (International ed. in English) 48(16):2983–2986.
Kondratenko E, Evgenii V, Pérez-Ramírez J (2006) Mechanism and kinetics of direct N2O
decomposition over Fe-MFI zeolites with different iron speciation from temporal
analysis of products. Journal of Physical Chemistry B 110(eq 3):22586–22595.
Law Y, Ye L, Pan Y, Yuan Z (2012) Nitrous oxide emissions from wastewater treatment
processes. Philosophical transactions of the Royal Society of London. Series B,
Biological sciences 367(1593):1265–1277.
Lebrero R, Rodríguez E, García-Encina P, Muñoz R (2011) A comparative assessment of
biofiltration and activated sludge diffusion for odour abatement. Journal of hazardous
materials 190(1-3):622–630.
Melse R, & Mosquera J (2014) Nitrous oxide (N2O) emissions from biotrickling filters used for
ammonia removal at livestock facilities. Water science and technology: a journal of the
International Association on Water Pollution Research, 69(5):994–1003.
Merchuk J, Ladwa N, Cameron A, Bulmer M, Pickett A, (1994) Concentric-tube airlift reactors:
effects of geometrical design on performance. AIChE Journal, 40(7):1105–1117.
Ravishankara A, Daniel J, Portmann R (2009) Nitrous oxide (N2O): the dominant ozone-
depleting substance emitted in the 21st century. Science (New York, N.Y.) 326:123–
125.
Richardson D, Felgate H, Watmough N, Thomson A, Baggs E (2009) Mitigating release of the
potent greenhouse gas N2O from the nitrogen cycle - could enzymic regulation hold
the key? Trends in biotechnology 27(7):388–97.
222
Tallec, G, Garnier J, Billen G, Gousailles M (2006) Nitrous oxide emissions from secondary
activated sludge in nitrifying conditions of urban wastewater treatment plants: effect of
oxygenation level. Water research 40(15):2972–2980.
Tallec G, Garnier J, Gousailles M (2006) Nitrogen removal in a wastewater treatment plant
through biofilters: nitrous oxide emissions during nitrification and denitrification.
Bioprocess and biosystems engineering 29(5-6):323–333.
Zheng M, He D, Ma T, Chen Q, Liu S, Ahmad M, Gui M, Ni J (2014) Reducing NO and N 2O
emission during aerobic denitrification by newly isolated Pseudomonas stutzeri PCN-
1. Bioresource technology, 162:80–88.
Zhu X, Chen Y, Chen H, Li X, Peng Y, Wang S (2013) Minimizing nitrous oxide in biological
nutrient removal from municipal wastewater by controlling copper ion concentrations.
Applied microbiology and biotechnology, 97(3):1325–1334.
223
Poster Contributions
Authors Title
Manuel Jesús Benito González, Isabel NEW SYNTHETIC CATALYST FOR TAR REFORMING IN BIOMASS AND
Ortiz González, José Maria Sánchez WASTE GASIFICATION STREAMS: COMPARATIVE STUDY vs. MINERAL
Hervás, Alberto Bahillo Ruiz MATERIALS AND COMMERTIAL CATALYST USING TOLUENE AS TAR
MODEL COMPOUND
José María Sánchez-Hervas, Virginia FLUIDISED BED GASIFICATION OF SWITCHGRASS, WHEAT TREE AND
Perez, José María Murillo, Yarima TOBACCO TREE FOR BIO-H2 PRODUCTION: PRELIMINARY RESULTS
Torreiro, Marta Maroño
Alma Toledo-Cervantes, Raúl Muñoz, BIOGAS UPGRADING IN CLOSED TUBULAR PHOTOBIOREACTOR
Tamara Morales, Raquel Lebrero
Patricio I. Cano, Javier Brito, Fernando SIMULTANEOUS AMMONIUN REMOVAL FROM AMMONIUN-RICH
Almenglo, F., Martín Ramírez, José M. WATER AND BIOGAS DESULFURIZATION
Gómez, and Domingo Cantero
Trisha L. Dupnock, Marc A. Deshusses UTILIZING HYDROGENOTROPHIC METHANOGENS IN A BIOTRICKLING
FILTER REACTOR TO UPGRADE RAW BIOGAS
I. Flores-Guerrero, L.J. Corzo-Rios and P AND N REMOVAL FROM WASTEWATERS COUPLED TO THE
L.G. Torres PRODUCTION OF BIOMASS AND LIPIDS USING A Chlorella CONSORTIUM
Javier Llorente, Sergio Díaz de Garayo, BIOMASS TRIGENERATION SYSTEM FOR RETAIL STORES
Daniel Zambrano
Luis Carlos Fernández-Linares, Mario Scenedesmus GROWTH IN 2 m3 RACEWAYS IN SEMICONTINUOUS
Andrés Bazán Peña and Mario Andrés REGIME FOR BIOFUELS PRODUCTION
Huante
Laura Vegas Magariño BIOENERGY IN AFRICA
Pietro Bartocci, Gianni Bidini, Giacomo THE ACTIVITIES ON PERENNIAL HERBACEOUS ENERGY CROPS IN THE
Tosti, Valeria Negri, L. Raggi, Franco BIOMASS RESEARCHE CENTRE, PERUGIA, ITALY
Cotana, Francesco Fantozzi
Ricardo Escalada Cuadrado, L. Alberto STUDY OF COMBUSTION KINETICS OF DIFFERENT WOODY BIOMASS BY
Bahillo Ruiz, Mª Elena Borjabad García, THERMOGRAVIMETRY
Raquel Ramos Casado
Y. Garcia, I. Garcia-Peña, L.G. Torres HETEROTROPHIC MICROALGAE CONSORTIUM GROWTH AND LIPID
PRODUCTION IN A WASTED MEDIA EMPLOYED FOR BIOHYDROGEN
PRODUCTION
Israel Díaz, Ivonne Figueroa-González, DETERMINATION OF THE OPTIMAL MIXING PROPORTION OF COW
Guillermo Quijano MANURE AND SLURRY TO MAXIMIZE BIOGAS PRODUCTION IN A DAIRY
FARM
Jorge Arenales Rivera, Ruth Barro SOLID RECOVERED FUELS: AN ATTEMPT OF CLASSIFICATION
Piñeiro, Raquel Ramos Casado
Mª Luisa Cosme Huertas LIFE MIX FERTILIZER
Pilar Lisbona, Yolanda Lara, Ana AMINE-IMPREGNATED SOLID SORBENTS CO2 CAPTURE:
Martínez, Luis Miguel Romeo HYDRODYNAMIC MODEL OF INTERCONNECTED FLUIDIZED BEDS
CONFIGURATION
Manuel Bailera, Pilar Lisbona, Luis M. REDUCING ENERGY PENALTY OF OXYCOMBUSTION THROUGH POWER-
Romeo, Sergio Espatolero TO-GAS HYBRIDIZATION
Silva-Rodriguez A., Sarabia-Martinez DESIGN OF A PARABOLIC CYLINDRICAL COLLECTOR FOR SOLAR
D.E., Vargas-Mondragon L.Z., Larios- THERMAL ENERGY APPLICATION
Ortíz E.A., Ruelas-Ramirez E.H.
Vargas-Mondragón L.Z., Sarabia- DEVELOPMENT OF CYLINDRICAL BI-PARABOLIC COLLECTOR FOR USE
Martínez D.E., Silva-Rodríguez A., IN THERMO SOLAR AND ITS EVALUATION WITH RESPECT TO THE CPC
Larios Ortiz E.A, Ruelas-Ramírez E.H. EQUIVALENT ENERGY
Keisy Torres, F. Javier Álvarez-Hornos, EFFECT OF CHITOSAN ON THE DEVELOPMENT OF GRANULAR
Pau San-Valero, Paula Marzal ANAEROBIC SLUDGE IN UPFLOW ANAEROBIC SLUDGE BLANKET
REACTOR TREATING VOLATILE ORGANIC COMPOUNDS
Pablo Ferrero-Aguar, Vicente MICROBIOLOGICAL EVOLUTION OF AN EGSB REACTOR TREATING
Martínez-Soria, Francisco Javier WASTEWATER CONTAINING INDUSTRIAL SOLVENTS
Álvarez-Hornos, Josep Manuel Penya-
roja
Alberto Ordaz, Ireri Robles INTENSIVE USE OF FIREWORKS DURING THE CELEBRATION OF THE
NEW YEAR IN THE VALLEY OF MEXICO
Estibaliz Diaz-Tena, Naiara Rojo, Ana DESIGN OF A BIOMACHINING PLANT FOR CONTINUOUS
Elías, Aida Aleksanyan, Gorka MICROETCHING
Gallastegui, Astrid Barona
224
NEW SYNTHETIC CATALYST FOR TAR REFORMING IN BIOMASS AND WASTE

GASIFICATION STREAMS: COMPARATIVE STUDY VS. MINERAL MATERIALS AND
COMMERTIAL CATALYST USING TOLUENE AS TAR MODEL COMPOUND
Manuel Jesús Benito González*, Isabel Ortiz González, José Maria Sánchez Hervás,
Alberto Bahillo Ruiz.
Energy Valorization of Fuels and Wastes Unit, Energy Department, CIEMAT, Av. Complutense 40,
28040 Madrid
*Corresponding author: mj.benito@ciemat.es, Tel. +34 913466701, Fax: +34 913466269.
Catalytic tar reforming is a potentially attractive process for tars removal in biomass and waste
gasification streams due to no additional energy input may be necessary, being the reaction
temperature range lower than gasification one. This fact minimizes the energy efficiency loss
in the overall process. A great number of materials have extensively studied in literature.
Among them, natural minerals such as dolomite, limestone and olivine are catalytically active
materials. However, disadvantages ascribed to its heterogeneous nature as well as the high
attrition leads to material losses and catalyst deactivation produced in most of the cases by
carbon deposition or re-carbonation when the partial pressure of CO2 is high.
From this standpoint, the design of new synthetic catalysts is the option to avoid the
deactivation phenomena increasing the catalytic activity, the bottleneck for the industrial
application of gasification processes. In this work a comparative reaction study on olivine,
commercial catalyst and a Ni-Co bimetallic catalyst synthetized in our group will be presented.
The catalysts were tested in a simulated gasification stream with a composition 11% H2, 7%
CO, 17,5% CO2, 4% CH4, 12,5% H2O and 48% N2 representative of sewage sludge air
gasification. Toluene was the tar model compound fed to the reactor in a concentration of 25
g·m-3, usually obtained in bubbling fluidized bed gasifiers. Two types of reaction tests were
performed: conversion vs. temperature and vs. time in order to study the catalytic activity and
stability. Opposite to olivine and commercial catalyst, the catalyst synthetized (CATAR3)
showed the lowest light-off temperature (Figure 1) of 494 oC while the light-off temperature for
olivine sample and commercial catalyst were 648 oC and 584 oC, respectively. Benzene was
the only by-product detected in the catalyst developed which is totally converted at
temperatures higher than 650 oC (Figure 2). The catalyst developed showed total stability
after 100 h reaction test without tar presence (Figure 3 and 4).
Keywords: Tar; Catalyst; Reforming; Gasification; Gas upgrading.
225
FLUIDISED BED GASIFICATION OF SWITCHGRASS, WHEAT TREE AND TOBACCO

TREE FOR BIO-H2 PRODUCTION: PRELIMINARY RESULTS
José María Sánchez-Hervasa, Virginia Perezb, José María Murillob, Yarima Torreiroa, Marta
Maroñoa
aCIEMAT, Unidad de Valorización Energética de Combustibles y Residuos, Avenida Complutense,
40, 28040 Madrid
bCEDER-CIEMAT, Unidad de Procesos de Conversión Térmica, Autovía A-15, salida 56, 42290 Lubia
(Soria) – España
*Corresponding author: josemaria.sanchez@ciemat.es, Tel. +34913466377, Fax: +34913466269
The main goal of the INNPACTO BIOH2 Project “New Strategies for Exploitation of Biomass
for Sustainable Production of Hydrogen without CO2 Release” MINECO, IPT-2012-0365-
120000 is to demonstrate the feasibility of producing renewable hydrogen to run a PEM fuel
cell and thus obtaining green energy. In order to do so, the project consists of different
workpackages covering the whole chain, from crops production, through feedstock
pelletisation, gasification, gas clean-up and upgrading, carbon dioxide capture and hydrogen
separation to finally electricity generation from the produced H2 in a PEM fuel cell.
In this work the results of the first gasification trials are presented and discussed. The three
selected biomasses, namely Panicum Virgatum (Switchgrass), Elytrigia Elongata (Wheat-tree)
and Nicotiana Glauca (Tobacco tree), were characterised and pelletised. Gasification
experiments were conducted in a pilot-scale bubbling fluidised bed gasifier. Olivine was used
as inert bed material. Tests were conducted using air, air+steam and air+oxygen as
gasification agents. Gas composition was measured on-line in order to determine the quality
of the producer gas. Moreover tar content was determined in accordance with CEN/TS
15439:2006. Biomass gasification – Tar and particles in producer gases.
Preliminary results show that it is feasible to gasify the three selected crops with little
differences in the composition of the gas produced. However, the gas needs further treatment,
including tar removal, upgrading by steam reforming and water shift reaction in order to
enhance hydrogen content and finally a hydrogen separation stage so that the gas can be fed
in a PEM fuel cell.
Keywords: Gasification, Biomass, Hydrogen, Fuel Cells, Energy production.
226
BIOGAS UPGRADING IN CLOSED TUBULAR PHOTOBIOREACTOR
Alma Toledo-Cervatntes, Raúl Muñoz*, Tamara Morales, Raquel Lebrero.
Department of Chemical Engineering and Environmental Technology, University of

Valladolid, Dr. Mergelina, s/n, 47011, Valladolid, Spain. Tel. +34 983186424
*
Corresponding author: mutora@iq.uva.es, Tel. +34 983186424.
Biogas production from the anaerobic digestion of organic residues in the European Union
has increased reaching 13.4 Mt of oil equivalent in 2013. This biogas can be used as a
renewable energy source for heat and electricity production in cogeneration plants, or as
substitute of natural gas after upgrading. Biogas typically contains 40-75% of CH4, 25-55% of
CO2, 0-2% of H2S, and other compounds at trace level concentrations. Removal of CO2 and
H2S from biogas is required to increase its energetic content and reduce transportation and
compression costs. Traditional physical-chemical processes are costly techniques, which
entail high energy and/or chemicals consumption rates. In this context, photobioreactors
based on algal-bacterial symbiosis constitute a cost-effective, environmentally friendly
technology capable of removing both contaminants from biogas.
A 110 L closed tubular photobioreactor (6 cm inner diameter) interconnected to a mixing
chamber and an absorption column (AC) was operated for biogas upgrading during 160 days.
A set of LED lamps providing an irradiance of ~1100 μmol m-2 s-1 was installed in order to
mimic sun irradiance. The synthetic biogas fed to the system was composed by 29.5% of CO2,
0.5% of H2S and 70% of CH4. The preliminary abiotic tests showed that a liquid to gas ratio
(L/G) of 11, allowed transferring ~91% and 98.6% of CO2 and H2S, respectively. Operation of
the photobioreactor at a HRT = 50 d, L/G = 11 and 12/12 h light/dark cycles resulted in
complete H2S removal, 99% of CO2 abatement and biomass productivity of 3.6 g d-1.
Keywords: Algal-bacterial technology, Biogas upgrading, CO2, Photobioreactors
227
SIMULTANEOUS AMMONIUN REMOVAL FROM AMMONIUN-RICH WATER AND

BIOGAS DESULFURIZATION
Patricio I. Cano*, Javier Brito, Fernando Almenglo, F., Martín Ramírez, José M. Gómez and
Domingo Cantero
Department of Chemical Engineering and Food Technologies. Faculty of Sciences. University of

Cadiz. 11510 Puerto Real (Cádiz), Spain. Campus de Excelencia Agroalimentario CeiA3.
*Corresponding author: patricio.cano@uca.es, Tel. +34956016340, Fax: +34956016411.
Hydrogen sulfide (H2S) is a common component that can be found in biogas streams. For
instance, in sewage treatments plants (STP) the H2S concentration can be up to 0.5% of the
biogas volume. Therefore, it is necessary to remove it to allow its use. Anoxic biotrickling filters
can achieve high elimination capacities (EC) up to 289.3 gS m–3 h–1 (EBRT: 164 s; RE:
99.35%) without biogas dilution. However it demands high amounts of nitrate, which is the
most important operational cost. Decreasing this cost could make anoxic desulfurization more
competitive for further scaling-ups.
At the same time, high ammonium (NH4+) concentration can be found some efluents of STP
such as the reject water from dewatering system of digested sludge. NH4+ could be removed
in a nitrification bioreactor, by oxidizing to nitrite (NO2–)/nitrate (NO3–) in a first stage and
carried out the denitrifying step in the anoxic BTF.
Therefore, verifying that there is not negative influence on anoxic desulfurization by a previous
NH4+ nitrification could improve both procedures. Firstly, there would not be necessary the
energy requirements to denitrify NO2-/NO3- and moreover, commercial NO3- usage would be
avoided.
Initially, it was tested that both procedures can be carried out without influence in long term
(45 days, IL 115.2 gS m–3 h–1, RE 96.5±2.9%) by a combined process: nitrifying bioreactor
(stirred tank) and anoxic BTF. Additionally, it was optimize the nutrients dosage, by decreasing
progressively the nutrients flow rate in the BTF up to 25% of the initial dosage, when it was
detected a drop in the removal efficiency (RE). It was carried out a step increasing H2S loading
rate experiment (from 57.5 to 289.3 gS m–3 h–1) using NO3- from the nitrifying bioreactor. The
results were compared with NO3- from chemical sources and it was obtained a similar
behavior. So, the both bioprocess combined allow to remove NH4+ from ammonium rich-water
and H2S from biogas reducing the operational costs.
Keywords: Hydrogen sulfide, Nitrification, Desulfurization, Biological integration, Cost

optimization.
228
UTILIZING HYDROGENOTROPHIC METHANOGENS IN A BIOTRICKLING

FILTER REACTOR TO UPGRADE RAW BIOGAS
Trisha L. Dupnock, Marc A. Deshusses
Department of Civil & Environmental Engineering, Duke University, Durham, North Carolina 27708
Upgrading raw biogas (~60% CH4, 40% CO2) produced via the anaerobic digestion of organic
wastes has emerged as a promising technology for promoting environmental sustainability. A
recent advancement that involves the addition of H2 and Hydrogenotrophic methanogens
(HMs) to the digester has been developed. While studies have conceptually validated that this
approach can produce CH4 effluent purities of up to 95%, low mass transfer rates of H2 from
the gas to liquid phases are severely limiting the volumetric production rates. Current
production rates are probably not economically feasible, thus it is critical to explore solutions
to the H2 mass transfer limitations.
This research investigates the use of biotrickling filters, which are known to have high mass
transfer coefficients, inoculated with HM’s to upgrade biogas at much higher rates. A bench-
scale bioreactor was packed with polyurethane foam and inoculated with HM’s enriched from
swine farm waste digestate. H2 and CO2 were fed to the bioreactor at a ratio of 60:40 vol.
based on the reaction stoichiometry along with inert N2, which served as an internal standard
for the quantification of the effluent flow. After a 12 day startup period, CH4 was detected in
the effluent gas. Maximum CH4 production rates have been recorded as high as 49 L
CH4/Lreactor-d, which is nearly five times faster than observed by others, coupled with H2 and
CO2 removal efficiencies of 73% and 90% respectively. Reactor pH is greatly dictated by the
amount of bicarbonate available in the liquid phase; as CO2 removal increases, pH increases
accordingly due to bicarbonate consumption. Further work involves conducting flasks
experiments with reactor biomass to determine HM kinetic parameters for both H2 and CO2
consumption, and conceptual modeling of the process. These all contribute to a better
understanding of how bicarbonate, kinetic, and transport limitations impact the bioreactor
performance.
Keywords: Biogas, Upgrade, Hydrogenotrophic methanogens, Biotrickling, Methane
229
P AND N REMOVAL FROM WASTEWATERS COUPLED TO THE PRODUCTION OF

BIOMASS AND LIPIDS USING A Chlorella CONSORTIUM
I Flores-Guerrero, LJ Corzo-Rios and L.G. Torres

1UPIBI-Instituto Politécnico Nacional. Bioprocess and Food departments. Acueducto s/n; Barrio
La Laguna Ticoman, 07342; Mexico City.
*Corresponding author: email: LGTorresBustillos@gmail.com, Tel 01+(55)57296000 ext 56367
Tertiary treatment of wastewaters for elimination of P and N, prevents the eutrophication

process in ponds, channels and natural reservoirs. This process can be carried out using
physicochemical and biological process. The use of microalgae growing autotrophically has
been reported because of its capacity of removing P and N, at the same time that an amount
of biomass of microalgae is obtained. This biomass can be used for lipids extraction followed
of the extraction of other interesting compounds (proteins, pigments, sugars, etc). Even
residual biomass can be employed for energy production, by m means of dark anaerobic
fermentation to produce methane. In that way and under the concept of biorefinery it is
possible to obtain various chemicals and energy.
In recent years, the use of dry baths is becoming popular. These artifacts consist of urinals (at
this stage only for gentleman) where urine is recollected without water addition. This urine
current can be employed with different purposes such as hydroponics, agricultural irrigation
and others. Some authors (Torres et al, 2014 among them) have demonstrated that urine in
water dilutions are very feasible to grow autotrophically microalgae, due to its high contents of
total nitrogen (1,100 mg/L) and P (170 mg/L as orthophosphates). In the same work, Torres
et al (2014) identified the appropriate dilutions for the culture of Chlorella vulgaris,
Scenedesmus sp. and Chlorococcum humicola. These dilutions were 1/50, 1/150 and 1/75,
respectively.
In this work, a consortium composed by Chlorella, Scenedesmus and some diatomaceous
was grown in urine dilutions (1/10 to 1/200) and the removal of N and P were evaluated, as
well as the biomass and lipids production. Using these data, P and N removal kinetics were
established. From these assessments it was shown that for the 1/200 dilution it was achieved
a N removal efficiency of 90%, and the lowest efficiency occurred with the 1/150 dilution (42%).
Regarding P, it was noted that for the 1/100 and 1/150 dilutions the best P removal efficiencies
were obtained (61%). At the same time, biomass concentrations up to 1,300 mg/L were
achieved. Lipids were produced at the different urine dilutions, but the final concentrations at
the end of the process were very similar among them, i.e. 5.87 mg/L.
At the end of the work, the same Chlorella consortium was used for the treatment of real
municipal wastewaters arising from the Plant located at San Juan Ixhuatepec (Edo de Mexico),
with excellent results.
Keywords: Biomass, Heterotrophy, Lipids, Microalgae consortia, P and N removal
230
BIOMASS TRIGENERATION SYSTEM FOR RETAIL STORES
Javier Llorentea*, Sergio Díaz de Garayob, Daniel Zambranoc

aCENER– Centro Nacional de Energías Renovables, Sarriguren, Navarra, Spain.
jllorente@cener.com
bCENER – Centro Nacional de Energías Renovables, Sarriguren, Navarra, Spain. sdiaz@cener.com
cCENER – Centro Nacional de Energías Renovables, Sarriguren, Navarra, Spain.
dzambrano@cener.com
*Corresponding author: jllorente@cener.com, Tel. +34 948 25 28 00, Fax: +34 948 27 07 74.
Retail stores are great contributors to global warming due basically to two factors: electricity
consumption and leak of hydrofluorocarbons to the atmosphere. The price of electricity
continues growing (4%) while new taxes are being applied for hydrofluorocarbons (planned
76 €/kg on average for 2016 in Spain). In this context new alternatives need to be put in place
to attend the energy demands in supermarkets.
This study analyzes the technological solutions planned for the energy refurbishement a 1600
sqm retail store in Vitoria (north of Spain) in the frame of the LIFE12 ENV/ES/000787 project.
The electricity consumption has been monitored during 2014 and 2015, and, based on the
results, a number of action packages have been designed.
The first action consists on energy demand reduction thanks to energy efficiency measures
on lighting and HVAC already installed and proven.
The second and core action consist on a trigeneration prototype based on biomass. A 250kWt
boiler will supply heat at 205ºC to an ORC, that willl produce electricity (27,5kWe) and residual
heat regulated at two levels (88ºC and 50ºC). 88ºC heat will power a Li-Br absoption chiller,
while 50ºC heat will be used for air heating or rejected when no necessary. The absorption
chiller (17,6 kWf) will be integrated with an indirect refrigeration cascade compression system
in order to feed medium and low temperature cabinets. Cascade system will be composed of
a N13 compressor (a new refrigerant with low GWP) and a CO2 compressor. The integration
with the absorption chiller is made by subcooling N13 after condensation. This way,
refrigeration will be ensured even if the prototype stops working.
This paper will describe this energy solution, analysing the energy perfomance simulated with
TRNSYS 17.0 based on real demand data. This study validates this novel 100% RES
trigeneration prototype.
Keywords: retail stores, energy efficiency, biomass trigeneration, absorption chiller, ORC.
231
Scenedesmus GROWTH IN 2 m3 RACEWAYS IN SEMICONTINUOUS REGIME FOR

BIOFUELS PRODUCTION
Luis Carlos Fernández-Linares*, Mario Andrés Bazán Peña and Mario Andrés Huante
Departamento de Bioprocesos, Unidad Profesional Interdisciplinaria de Biotecnología, Instituto
Politécnico Nacional, México.
*Corresponding author: lfernand36@hotmail.com. Tel. +52 57296000 Ext. 56387
Microalgae have the ability to synthesize precursor molecules for the potential production of
renewable fuels like bioethanol, bio-hydrogen, bio-methane or biodiesel. Microalgae can be
grown in open Raceways (RW). The main cultivation modes in RW are the batch, feed-batch
and cyclic fed-batch (CFB). The CFB consists of periodic withdrawals of culture coupled to
constant medium feeding; CFB allows to maintain log phase growth; avoids the shading effect;
and allows inoculum availability during cycle.
In this work, the effect of cyclic fed-batch culture of Scenedesmus on lipids and biomass
production was studied in 2000-L raceway ponds during 9 harvesting cycles under
greenhouse conditions. The highest biomass productivity was 53.1 mg/Ld on days 45 (3th
cycle). The overall range of biomass productivity was between 17.7 and 8.22 mg/Ld. Respect
to the productivities of sugars, proteins and pigments the overall range were 19.2-5.6, 2.9-0.8,
and 1.4-3.9 mg/Ld, respectively. The CFB culture of Scenedesmus in RW showed an
increasing biomass production in each new cycle during the first 4 cycles; then a biomass
reduction was observed. The RW minimum and maximum temperature range was from 13 to
18 and from 19 and 31.6 C, respectively. With maximum irradiance in the day between 48230
to 67560.
Average biomass productivity was 43.87 mg / dL and decreased by 40% after the third cycle.
The maximum levels of protein, sugars and lipids were 5.58%, 41.5% and 38.4% dry weight,
respectively. Weather conditions are very important in the productivity and the biochemical
composition of Scenedesmus sp.
Keywords: Microalgae, Scenedesmus, Raceways, Biofuels
232
BIOENERGY IN AFRICA
Laura Vegas Magariño*
Independent consultant Renewable Energy Africa, Toreros 21, 47007 Valladolid, Spain
*Corresponding author: laura.vegmag@gmail.com, Tel. +34 600202349
Energy access around the globe is now an issue in the international agenda. Launched in
2011 by Ban Ki-moon, the “Sustainable Energy for All” Initiative pretends to achieve by 2030:
i) universal access to modern energy services; ii) double the global rate of improvement in
energy efficiency; iii) double the share of renewable energy in the global energy mix
(www.se4all.org). The Africa-EU Energy Partnership is, from December 2007, a long-term
framework for dialogue and cooperation between Africa and Europe in energy related issues
(AEEP, 2011). In 2012 the United Nations General Assembly declared 2014-2024 as the
Decade of Sustainable Energy for All (www.un.org). Sustainable Development Goals, signed
in September 2015 by more than 150 leaders in the framework of the 2030 Agenda for
Sustainable Development, include SDG7: Affordable and clean energy (www.undp.org).
Renewable energy sources are considered by different civil and non-governmental
organisations as a powerful solution to poverty (www.practicalaction.org).
Solid biomass is the most used energy source in Sub-Saharan Africa - 70% of final energy
consumption (IEA, 2014). Harmful effects have been pointed out from inefficient firewood
combustion, air pollution (Smith, 2006), child burns, increased time and insecurity for picking
up wood, deforestation (IEA, 2006). Locally, land competition with local communities has to
be avoided (Renewable Fuels Agency UK, 2008; Practical Action Consulting/ACTS, 2011)
and sustainable technologies need to be implemented (www.ecreee.org).
Bioenergy is expected to develop a role in the future African energy balance (IEA, 2014),
contributing to national energy security and emissions mitigation, and supplying additional
income to rural populations. Technologies are being developed and implemented in
SubSaharan Africa remote communities for sustainable energy access (improved cookstove
combustion, biomass gasification, biochar and biogas production, anaerobic digestion,
biodiesel production connected to agricultural businesses, off-line mini-grids). A
comprehensive analysis of energy, agricultural, forestry, science and technology, social and
development aspects will be provided with a Social Sustainable Energy Approach
(https://africa-energy.org/). A review on successful technologies adapted to the African context
and on operating bioenergy projects will be made available.
“Tree planting is simple, attainable and guarantees quick, successful results within a
reasonable amount time” (Wangari Maathai, 2004).
Keywords: Biomass; Biofuels; Africa; Energy access; Bioenergy technologies
233
THE ACTIVITIES ON PERENNIAL HERBACEOUS ENERGY CROPS IN THE BIOMASS

RESEARCHE CENTRE, PERUGIA, ITALY
Pietro Bartocci*a, Gianni Bidinia, Giacomo Tostib, Valeria Negrib, L. Raggib, Franco Cotanaa,
Francesco Fantozzia
aDepartment of Engineering, University of Perugia, Perugia, Italy
b Department of Agronomics and Food and Environmental Sciences, Univ. of Perugia, Borgo XX
Giugno 74, Perugia, Italy
*Corresponding author: bartocci@crbnet.it, Tel. +39 0755853773, Fax: +39 0755153321
To feed big bioenergy plants such as bio-refineries millions of tons of biomass per year are
required in Europe. These can be produced with efficient herbaceous C4 crops, which
cultivation can be easily mechanized and managed by farmers. The growing biobased industry
needs to produce its feedstock through the cultivation of perennial herbaceous crops. For this
reason the Biomass Research Centre (University of Perugia, Italy) has tested since 2007
different crops, such as: miscanthus, giant reed, cardoon and switchgrass to assess their
productivity, to optimize their agronomic management and to evaluate the energy yields. In
this way the foundation for future genetic improvement of the actual varieties can be layed.
This paper presents the most important results on the experimentation on the above
mentioned crops both in terms of productivity and energy yields. Besides further insights on
agronomic and genetic aspects of energy crops are provided through the analysis of
researchers of the Department of Agronomics and Food and Environmental Sciences of the
University of Perugia.
Keywords: Biomass; Switchgrass; Miscanthus; Cardoon; Giant reed
234
STUDY OF COMBUSTION KINETICS OF DIFFERENT WOODY BIOMASS BY

THERMOGRAVIMETRY
Ricardo Escalada Cuadradoa, L. Alberto Bahillo Ruizb, Mª Elena Borjabad García a Raquel
Ramos Casadoa
aCEDER-CIEMAT Autovía Navarra A15, salida 56, 42290 Lubia (Soria), Spain.
bCIEMAT, Avda. Complutense 40, 28040 Madrid, Spain
*Corresponding author: ricardo.escalada@ciemat.es, Tel. +34 975281013+208, Fax: +34 975281051.
Combustion is the main process to obtain energy from biomass. Many residual woody biomass
are available for energy valorisation. Lignocellulosic biomass is a matter of very complex
chemical composition, consisting mainly of polymers: hemicellulose, cellulose and lignin. The
thermally stimulated oxidative decomposition and its kinetics is a very interesting issue for
modelling thermal conversion processes.
The macroscopic kinetics is inherently complex because it includes information about multiple
steps that occur simultaneously. As indicated in year 2000 by the results of the ICTAC Kinetics
Project, only the methods that use multiple heating programs can meet this challenge.
The development of models with kinetics parameters is useful to understand the behaviour of
fuels under combustion or gasification processes, which occur at high temperatures. The
common starting point is the general rate equation, usually expressed as
= 𝑘(𝑇) · 𝑓(𝛼) = 𝐴 · exp (−𝐸(𝛼)⁄𝑅 · 𝑇) · 𝑓(𝛼)
𝑑𝛼
𝑑𝑡
Ec. 1
where k(T) is the temperature function, f() is the conversion function and E() is the activity
energy function. But Ec. 1 is only a rough approximation of the actual kinetics.
Data for E() of four different woody biomass are compared in this paper. Some traditional
isoconversional methods (Friedman, FWO, KAS) and other modern methods, that are more
accurate, (incremental flexible integral methods) are used.
Finally, bibliography data of hemicellulose, cellulose and lignin kinetics are integrated in a
model and compared with experimental results.
Keywords: Biomass; Kinetics model; Isoconversional method; Thermogravimetry.
235
HETEROTROPHIC MICROALGAE CONSORTIUM GROWTH AND LIPID PRODUCTION

IN A WASTED MEDIA EMPLOYED FOR BIOHYDROGEN PRODUCTION
Y. Garcia, I. Garcia-Peña, L.G. Torres

UPIBI-Instituto Politécnico Nacional. Bioprocess department. Acueducto s/n; Barrio La Laguna
Ticoman, 07342; Mexico City.
*Corresponding author: email: LGTorresBustillos@gmail.com, Tel 01+(55)57296000 ext 56367
Biohydrogen is a very interesting alternative at present, because it is a clean and high energy
density biofuel. One of the ways which it can be produced efficiently is the dark fermentation
of wastes (such as urban municipal solids). The first stage of acidogenic dark fermentation is
an acidogenic step, where complex molecules rise to a large number of simple molecules such
as volatile organic acids AGV. In this stage, high concentrations (in the order of g/L) of acetate,
propionate, butyrate and lactate are achieved. In a second stage, part of those AGV is
converted to biohydrogen by the existing microflora. The remaining AGV, are left in the
fermentation broth (Gomez-Romero et al, 2014). It has been reported (Turon et al, 2015) that
some microalgae can grow heterotrophically in AGV mixtures and biomass can be dried so
lipids can be extracted and the residual materials can be converted to CH4 via dark
fermentation process. In that way, the higher amount of energy can be obtained (H2, biodiesel,
CH4 and heat).
In this work the production of biomass and the generation of total lipids using the waste media
arising from the dark fermentation for hydrogen production by means of a consortium
containing Chlorella, Scenedesmus and diatomaceas is evaluated. The consortium was able
to grow on acetate, propionate, butyrate and lactate at maximum concentrations of 2, 1, 1 and
5 g/L, respectively. Afterwards, the consortium was grown in two and three AGV mixtures. It
was capable of growing in a mixture acetate-propionate butyrate (1, 1 and 0.25 g/L,
respectively) and reach 0.65 g/L of biomass. The consortium was able to grow in the undiluted
medium used for fermentation production of biohydrogen removing up to 71% of the initial
COD (960 mg/L). The process was evaluated at the media dilution that succeed in the previous
step (no dilution) changing the pH of the initial media (6, 7, 8, 9 and 10). The consortia was
able to growth at various pH values, but best results were obtained at pH=7, the natural pH of
the wasted medium. These findings suggest the feasibility of producing massively biomass,
lipids and even other products using the wasted media generated in the biohydrogen
production.
Keywords: Heterotrophy, Acetate, Butyrate, Microalgae consortia, Propionate, Waste media
236
DETERMINATION OF THE OPTIMAL MIXING PROPORTION OF COW MANURE AND

SLURRY TO MAXIMIZE BIOGAS PRODUCTION IN A DAIRY FARM
Israel Díaza*, Ivonne Figueroa-Gonzáleza,b, Guillermo Quijanoa,b

a Department of Chemical Engineering and Environmental Technology, Escuela de Ingenierías
Industriales, Sede Dr. Mergelina, University of Valladolid, Dr. Mergelina s/n, 47011 Valladolid, Spain
bAgrarian Engineering School – University of Valladolid, Soria Campus. 42004 Soria, Spain.
*Corresponding author: israel.diaz@iq.uva.es, Tel. +34 983186424
The characteristics of cow manure and slurry in dairy farms strongly depend on farmer’s
practices. In general, manure presents a high solid content and low anaerobic biodegradability
when straw is employed as bedding, while slurry may achieve a low C/N ratio that results in
poor biogas yield. The simultaneous digestion can increase the C/N ratio to the optimal for AD
and increase biogas production but the low digestibility of straw may result counter-productive
if added in an excessive proportion. In this study, the influence of the combined digestion of
both residues from a 1000 cows dairy farm was assessed and an energetic balance was
performed. For this purpose, the digestion of cow slurry combined with 10%(w.) and 20%(w.)
of cow manure was compared with the digestion of cow slurry alone in BMP tests for 62 d.
Cow slurry alone and cow slurry with 10% manure showed a similar productivity around 190
LCH4 / gVS (NC)after 16 d; on the contrary, cow slurry achieved only 80 LCH4 / gVS after the same
period. Then, the digestion of cow slurry with 10%(w.) manure was the most convenient
mixture to maximize biogas production since the amount of residue to digest was only
increased by 11.1% while the methane production rose by 30% from 8.4 LCH4 / gsample to 11.5
LCH4 / gsample. As a consequence, a slightly larger digester to assume the digestion of 10%
manure with slurry results in a considerably larger electricity and heat recovery than the
digestion of slurry alone. For the optimal mixing conditions, an energetic balance for the
implementation of the AD in the farm was performed. AD of manure would produce 1330
MWh/yr while the energy consumption of the farm and the AD plant would be 749MWh/yr,
thus generating a surplus of 580MWh/yr.
Keywords: BMP, Dairy, Manure digestion, Manure valorization, Digestion optimization.
237
SOLID RECOVERED FUELS FOR ENERGY RECOVERY
Jorge Arenales Rivera*, Ruth Barro Piñeiro, Raquel Ramos Casado
CEDER-CIEMAT, Autovía de Navarra A15, Salida 56, 42290 Lubia (Soria), Spain
*Corresponding
author: jorgearenales@yahoo.es, Tel. +34 975281013, Fax: +34975281051.
The European legislation about waste disposal has established a hierarchy of treatment that
imposes a priority order for waste management: prevention, preparing for re-use, recycling,
recovery and disposal. Nowadays, different options for energy recovery such as biological
(anaerobic digestion) or thermal treatment (combustion, gasification, etc.) are available. The
interest of the European industry, generally cement plants, in finding alternative energy
sources to fossil fuels has led to the development of a series of European standards for Solid
Recovered Fuels (SRFs) production, sampling, analysis and characterization. According to
EN 15357:2011, SRFs are solid fuels prepared from non-hazardous waste to be used for
energy recovery in incineration or co-incineration plants and meeting the classification and
specification requirements laid down in EN 15359:2011. This classification system is based
on three important fuel characteristics, which can be considered as indicators for economic
(net calorific value), operational (chlorine content) and environmental (mercury content)
issues. The aim of this study is to assess the suitability of different wastes to be used as raw
materials for secondary fuel production. The wastes analyzed were: paper industry wastes
and municipal solid wastes (MSW). The selected streams in each treatment plant were
processed and compressed to obtain a homogeneous product, suitable to be used as
feedstock in energy facilities. The produced SRFs have been characterized in an attempt to
discern which SRF class these secondary fuels belong to. When mixing the paper industry
wastes, four secondary fuels were obtained by blending of two different waste fractions (paper
rejects and de-inking sludge), whose classification based on the three properties were: NCV
2 (24.11 MJ/kg), Cl 5 (1.56%) and Hg 1 (0.002 mg/MJ) for mixture 1; NCV 2 (21.89 MJ/kg), Cl
5 (1.55%) and Hg 1 (0.002 mg/MJ) for mixture 2; NCV 3 (19.19 MJ/kg), Cl 4 (1.45%) and Hg
1 (0.002 mg/MJ) for mixture 3; NCV 3 (18.49 MJ/kg), Cl 4 (1.17%) and Hg 1 (0.004 mg/MJ)
for mixture 4. Regarding MSW, four different waste fractions (rejects from Mechanical
Biological Treatment plant (MBT), rejects from packaging, mixed plastic and film) were studied
to use a raw materials for secondary fuel production, whose classification was NCV 2 (20.34
MJ/kg); Cl 3 (0.76%) and Hg 1 (0.005 mg/MJ). In addition to the classification system, this
standard contains a specification system based on properties that are obligatory and non-
obligatory to specify. Hence, composition, origin and preparation, biomass content, physical
parameters (bulk density, ash melting behavior), chemical parameters (major and trace
elements) were determined for these fuels.
Keywords: Solid Recovered Fuel; Refuse Derived Fuel, Paper Industry Wastes, Residual
Municipal Solid Waste; Characterization
238
LIFE MIX FERTILIZER
Mª Luisa Cosme Huertas

Oficina Técnica I+D+i Purines Almazán, S.L. Centro de Transferencia de Tecnología
Aplicadas. Oficina 106 Pº Belén, 9A. 47011 Valladolid (España)
This project aims to develop an innovative system for the treatment of the waste from the
anaerobic digestion of pig manure (digestate) for the development of a new type of fertilizers
with an organic/mineral base and gradual release of the nitrogen.
The solid fraction of the digestate obtained at the biogas plant of Pural has been composted
with other organic wastes such as chicken manure and vegetable biomass. The liquid fraction
of the digestate is being treated in a stripping prototype for the recovery of the nitrogen as
ammonium sulphate.
Servimed has produced a new type of fertilizers by the combination of the organic fraction and
the ammonium sulphate previously obtained and a nitrification inhibitor to control the nitrogen
release. ASAJA-Soria is evaluating the effectiveness of these new fertilizers in controlled
barley rainfed trials. Because of their dual nature, the new fertilizers will have a number of
environmental and economic benefits over currently marketed products such as soil quality
improvement, minimization of nitrogen losses through nitrate leaching, GHG emissions
reduction, agricultural yields increase and cost reduction by a single application.
Keywords: Renewable energy; Collector’s cylinder parabolic; Solar energy
239
AMINE-IMPREGNATED SOLID SORBENTS CO2 CAPTURE: HYDRODYNAMIC MODEL

OF INTERCONNECTED FLUIDIZED BEDS CONFIGURATION
Pilar Lisbonaa*, Yolanda Larab, Ana Martínezb, Luis Miguel Romeoc

a Escuela Universitaria de Ingenierías Agrarias de Soria – Universidad de Valladolid,
Campus Universitario Duques de Soria, 42004, Soria, Spain
b Research Centre for Energy Resources and Consumption (CIRCE) – Universidad de Zaragoza,
CIRCE Building – Campus Río Ebro, Mariano Esquillor Gómez, 15, 50018, Zaragoza, Spain.
c Department of Mechanical Engineering – Universidad de Zaragoza – Campus Río Ebro, María de
Luna 3, 50018, Zaragoza, Spain.

*Corresponding author: mariapilar.lisbona@uva.es, Tel. +975 129418
Amine-impregnated Alumina Solid Sorbent Carbon Capture (ASC2) is a very interesting option
to improve carbon capture efficiency and reduce the energy penalty in the power plant
associated to the carbon capture cycle. The regeneration stage of these sorbents requires
less energy due to absence of water and the high heat capacity of solids. Configurations which
include two interconnected circulating fluidized beds are promising for continuous operation.
Up to the moment, lab scale hydrodynamic feasibility of the amine-impregnated alumina CO2
capture technology has not been checked. The required design to make this technology
applicable at industrial scale have been investigated within the AMICO2 project and presented
in this work. Carbonation conditions of these sorbents make the lab-scale cold flow facility
quite different from other well-known solid looping processes when it is designed based on
Glicksman’s scaling rules. The design of the fluidized bed configuration and the solid material
selection, fluidizing agent and cold-flow operating conditions must resemble the circulation of
solids between reactors and ensure sufficient solids inventory in the reactors to promote high
absorption/desorption efficiencies.
Experimental tests and mathematical modelling are required to establish a comprehensive

control method to operate the interconnected fluidized beds in their corresponding
hydrodynamic regime. Measurements of circulation rate, static pressure, voidage profiles and
standpipe height of solids have been planned to identify trends in the hydrodynamic behaviour
of the whole system while varying gas velocities, loop-seals, inventories in the reactors or
particles size distribution.
Keywords: Hydrodynamics, amine-impregnated solid sorbent, carbon capture, fluidized beds.
240
REDUCING ENERGY PENALTY OF OXYCOMBUSTION THROUGH POWER-TO-GAS

HYBRIDIZATION
Manuel Baileraa*, Pilar Lisbonab, Luis M. Romeoa, Sergio Espatoleroa

aResearch Centre for Energy Resources and Consumption (CIRCE), Universidad de Zaragoza,
CIRCE Building, Campus Río Ebro, Mariano Esquillor Gómez, 15, 50018 Zaragoza, Spain
bEscuela Universitaria de Ingenierías Agrarias de Soria, Universidad de Valladolid, Campus
Universitario Duques de Soria, 42004 Soria, Spain

*Corresponding author: mbailera@fcirce.es, Tel. +34 976761863, Fax: +34 976732078.
The constant increase of electricity production from renewable energy has brought to light the
necessity of deploying energy storage systems. The management of intermittent surplus
power generated by renewable sources is crucial for reaching a future reliable electricity
system. Power-to-gas (PtG) has been proposed as a promising solution to overcome this
problem.
PtG stores electricity in natural gas form through catalytic methanation of hydrogen from
electrolysis. This technology broadens the application of hydrogen as energy vector by
connecting the electric and gas networks as a unique system thus increasing the supply
flexibility. Furthermore, synthetic natural gas from methanation is “CO2 neutral” since uses
carbon emissions.
In order to take advantage of the oxygen generated by electrolysis, it is proposed the

hybridization with oxyfuel combustion. This carbon capture method uses a O2/CO2 mixture as
comburent to burn the solid fuel which generates flue gases mainly composed by CO2 and
water (easily removed from the mixture). Highly concentrated CO2 is obtained avoiding the air
separation unit for oxycombustion, thus reducing the associated energy penalty an 81.4%.
The key variable in this system is the size ratio relating the energy contained in the hydrogen
produced by electrolysis and the useful thermal output of the oxyfuel boiler. One of the most
suitable applications of the hybridization is district heating (1-2MW), in which up to the 81% of
thermal energy from methanation could be used.
Keywords: Power-to-Gas; methanation; SNG; Oxycombustion.
241
DESIGN OF A PARABOLIC CYLINDRICAL COLLECTOR FOR SOLAR THERMAL

ENERGY APPLICATION
Silva-Rodriguez, Aa., Sarabia-Martinez, D.E b. Vargas-Mondragon, L.Z.c, Larios-Ortíz, E.A.d,

Q.F.B. Ruelas-Ramirez, E.H.e
Universidad Mexiquense del Bicentenario, Unidad de Estudios Superiores Tultitlan, Av. Ex Hacienda
de portales s/n Col. Villa esmeralda, Tultitlan, Estado de México, C.P. 54910, México.
lucero.z.v,m@gmail.com, diana.sarabia87@gmail.com, Analilia.Silva.R@gmail.com ,
Edwinaloing@gmail.com eh.ruelas@umb.mx.
The limited supply of fossil fuels and the negative impact of the emission of carbon dioxide on
the global climate change dictate the increasing use of renewable energy sources. Inside the
renewable energy sources there is the wind power, the marine one, the thermal solar energy,
etc., the thermal solar energy stands out for being a viable and feasible technology. Since
1990, renewable energy sources have grown at an average annual rate of 1.7 %, which is
slightly lower than the growth rate of world oil equivalent tons of 1.9 % per year. The countries
of the Organisation for Economic Co-operation and Development (OECD) account for most of
production and the growth of solar and wind energy. Solar photovoltaic and solar thermal
recorded an annual growth rate of 9.8 %. The solar energy has different applications
technologies commercially available as sound those toasts lot or a system of central recipient,
cylinder collector’s parabolic. The collectors play a main role since they can reach large
numbers of solar radiation, this one concentrates on a small area. The different efficiency of
the collectors, depending the type of collector (form designed in general, material, approach
and size).
Nevertheless most of these parabolic collectors have a big size and a low incidence level, this
technology has showed an economic increase in the materials with which they are made,
since they take some recovering in silver or platinum, the above mentioned material has a
good thermal conductivity but it goes so far as to have a high cost. This is the reason by which
it is looked obtaining a better use of the incidence of the solar radiation and power to improve
the costs in this technology, to make her profitable for applications in different chemical
processes. The use of this technology will allow improving the valuation of energy comeback
(TRE) of different processes contributing to the process development sustainable. The primary
objective of this research is the construction of a parabolic low-cost cylindrical collector made
mainly with cheap materials. To meet this goal, a bibliographic search was done to find out
the materials meeting the requested characteristics. The result of this search yielded materials
having good thermal conductivity such as steel, polycarbonate sheet, the aluminum, copper,
alu foil, etc. Cost evaluation of the above mentioned materials showed that polycarbonate and
the reflective film has a high thermal conductivity and resulted a good material for construction
of a cylidrical parabolic collector.
Keywords: Renewable energy; Collector’s cylinder parabolic; Solar energy
242
DEVELOPMENT OF CYLINDRICAL BI-PARABOLIC COLLECTOR FOR USE IN THERMO

SOLAR AND ITS EVALUATION WITH RESPECT TO THE CPC EQUIVALENT ENERGY
Vargas-Mondragón, L.Z., Sarabia-Martínez, D.E., Silva-Rodríguez, A., Larios Ortiz, E.A,

Q.F.B. Ruelas-Ramírez, E.H.
Universidad Mexiquense del Bicentenario, Unidad de Estudios Superiores Tultitlan, Av. Ex Hacienda
de portales s/n Col. Villa esmeralda, Tultitlan, Estado de México, C.P. 54910, México.
lucero.z.v,m@gmail.com, diana.sarabia87@gmail.com, Analilia.Silva.R@gmail.com ,
Edwinaloing@gmail.com eh.ruelas@umb.mx.
Today one of the priorities is the continued development of alternative energies, to help
replace electricity thus having a clean environment. Energy is an essential part for the
development of societies. Subject to scientific advances in renewable energies (solar, wind,
biomass. Etc.) as they represent an alternative that consideration be able to processes that
need some kind of energy. Alternative energy sources currently being investigated. Solar
energy is a viable option to replace power sources in different applications option, since in just
one hour, the sun transmits more energy to Earth than is consumed in a year. This is the
reason why solar energy will be one of the main pillars for energy production in the future.
Beside the use of solar energy to produce heat (thermal solar energy), the sun will also be
used to generate electrical energy, which is also known as solar photovoltaic (PV). Taking
advantage of solar radiation properly, we can obtain heat and electricity. However the current
designs of technologies used for the use of such energy, such as CPC and solar cells have
an efficiency of only 60% so do not fail 40% energy and is wasted, adding to this poor existing
technology causing diversity of applications is lowered because only is used to heat water,
support as solar cooking among others, and also these systems are large to achieve the
desired temperatures having low solar efficiency, this is caused the design of these systems,
coming from the size to its structure. The aim of this work is the development of a new design
of CCP in order to significantly improve efficiency in the incidence and capacity of collecting
solar thermal energy to transform it into electricity, plus the design has better targeting and
focus rays.
The bi-cylindrical parabolic collector (CCBP) has a structure of two parabolas with a single
focus receiver having a 90 ° angle so that the rays are parallel to the parabolic channel and
so are bounced continuously to the receiver tube, the efficiency achieved with a first design is
70%, the project aims to further modify to considerably improve the efficiency of these
systems, by the time the first design prototype has thrown us promising results, and that the
use of two parables in one system as address again rays at different points receptor improves
the efficiency of solar thermal energy captured.
Keywords: Solar thermal energy, Collector parabolic, Technologies, Efficiency, Solar

radiation
243
EFFECT OF CHITOSAN ON THE DEVELOPMENT OF GRANULAR ANAEROBIC

SLUDGE IN UPFLOW ANAEROBIC SLUDGE BLANKET REACTOR TREATING
VOLATILE ORGANIC COMPOUNDS
Keisy Torres*, F. Javier Álvarez-Hornos, Pau San-Valero, Paula Marzal.
Research group GI2AM, Department of Chemical Engineering, University of Valencia, Avda.

Universitat s/n, 46100 Burjassot, Spain*
*Corresponding author: Keisy.Torres@uv.es, Tel. +34 963544996, Fax: +34963544898
In the present study, the influence of chitosan polymer on the development of granular
anaerobic sludge in upflow sludge blanket reactor (UASB) treating a mixture of volatile organic
compounds is evaluated. The experimental set up consists of three laboratory-scale UASB
reactors with a reaction volume of 4.0 liters (internal diameter of 65 mm and height of 1200
mm) and an overall volume of 7.8 L, fed with synthetic wastewater composed of ethanol, ethyl
acetate and 1-ethoxy-2-propanol in a mass ratio of 3.5:1:0.5. A control reactor R1 is operated
without chitosan addition, while the other two reactors (R2 and R3) are operated at the same
conditions with a single chitosan dosage applied on the first operating day. The optimal dosage
was determined in a previous flocculation test using a suspended sludge from an anaerobic
digester of a municipal wastewater treatment plant, which corresponds to the seed sludge of
the reactors. Chitosan dosages of 0.6, 1.0, 1.4, 2.0, 2.4, 3.0 and 4.0 mg/gVSS were evaluated.
Approximately 90% flocculation was obtained with a chitosan concentration of 2.4 mg/gVSS.
Currently, the investigation is at the startup period. The reactors are operated at an initial
organic load rate (OLR) of 2.25 kg COD/L.d with COD removal efficiency of 85% for R1 and
90% for R2 and R3. Sludge characteristics such as sludge volume index (SVI), particle size,
settling velocity and specific methanogenic activity are evaluated, as well reactors
performance under the influence of the polymer and mixture substrate.
Keywords: Granulation; UASB; Volatile organic compound; Chitosan.
Acknowledgments: Authors gratefully acknowledge the financial support by Ministerio de Economía

y Competitividad (Project CTM2014-54517-R) and Generalitat Valenciana (PROMETEO/2013/053),
Spain. Keisy Torres acknowledges the Generalitat Valenciana for the Grisolia Grant
(GRISOLIA/2015/A/021).
244
MICROBIOLOGICAL EVOLUTION OF AN EGSB REACTOR TREATING WASTEWATER

CONTAINING INDUSTRIAL SOLVENTS
Pablo Ferrero-Aguar *, Vicente Martínez-Soria, F. Javier Álvarez-Hornos, Josep Manuel

Penya-roja
Research group GI2AM, Department of Chemical Engineering, University of Valencia, Avda.

Universitat s/n, 46100 Burjassot, Spain *
Corresponding author: Pablo.Ferrero@uv.es, Tel. +34 963544996, Fax: +34 963544898.
The aim of this research was to study the microbiological evolution during more than 20
months of an anaerobic expanded granular sludge bed (EGSB) reactor treating wastewater
containing typical industrial solvents used in the food packaging as ethanol, ethyl acetate and
1-ethoxy-2-propanol to produce biogas. The bioreactor was seeded with anaerobic granular
sludge from a brewery wastewater treatment plant and was operated at organic loads ranging
from to 17 to 38 kg COD m-3 day-1 with removal efficiencies above 90%. Denaturing gradient
gel electrophoresis (DGGE) technique was selected and set up to monitor the archaeal and
bacterial populations in the bioreactor. DGGE bands were cut and sequenced to identify the
predominant microorganisms. The results of the archaeal DGGE pointed out the
establishment of a stable community with a relatively low-diversity (Methanosaeta concilii, as
predominant specie), which was developed from an inoculum with a much higher biodiversity.
Regarding the bacterial DGGE results, a clearly defined evolution during bioreactor operation
was observed, since intensity of some initial predominant bands decreased with operation
time, even disappearing, and other bands increased their intensity, and even some of them
(identified as belongings to Geobacter species) became predominant. This study discloses
that the main microorganisms involved in this process to obtain bioenergy are Methanosaeta
and Geobacter species, and shows DGGE as a suitable method to monitor and identify
microbiological changes in these bioreactors, which could be useful to understand, and even
control, their performance.
Keywords: anaerobic reactor, DGGE, biomethane; Methanosaeta; Geobacter.
Acknowledgments: Authors gratefully acknowledge the financial support by Ministerio de Economía

y Competitividad (Project CTM2014-54517-R co-financed with FEDER funds) and Generalitat
Valenciana (PROMETEO/2013/053), Spain. Pablo Ferrero-Aguar acknowledges the Ministerio de
Economía y Competitividad for the FPI contract (BES-2015-073319).
245
INTENSIVE USE OF FIREWORKS DURING THE CELEBRATION OF THE NEW YEAR IN

THE VALLEY OF MEXICO
Alberto Ordaz and Ireri Robles
Unidad de Estudios Superiores de Tultitlán, Universidad Mexiquense del Bicentenario, Av. Ex-
Hacienda de Portales s/n, Villa Esmeralda, Tultitlán de Mariano Escobedo, México, 54910, México.
Corresponding author: alberto.ordaz@umb.mx
Around the world, fireworks displays are widely used for religious, cultural, sports, national
and family celebrations; although those are recreational and entertainment activities, it has
caused concern about the emission of pollutants that harm not only air quality but may produce
toxic effects on the ecosystems and people attending those events. Particularly it has been
studied that the massive burning of fireworks generates an increase in the level of atmospheric
particulate matter (PM 2.5 and PM 10) in places close to the fireworks displays, the typical
composition of the pollutants include metals (such as Na, K, Sc, Cr, Mn, Fe , Co, Zn, as, Br,
Sr, Sb, Pb, Cs, Ba, Ti, V and Cu) and some anions such as perchlorate, sulfate and nitrate;
also, greenhouse gases are generated (ozone, CO2 and NOx). In Mexico, particularly in the
State of Mexico it is produced about 50% of the country fireworks, and only the town of
Tultepec concentrates most of the government
official permits to produce fireworks. In the urban area of the Valley of Mexico and during the
first hours of the New Year, high concentrations of particulate matter are recorded, which are
mainly attributed to the massive burning of fireworks. The purpose of this document is to make
a comprehensive review of the pollutants emitted during the displays of fireworks at various
festivals around the world and warning about this type of pollution that has gone unnoticed in
the scientific community of Mexico.
Keywords: Fireworks, Heavy metals, Mexico City
246
DESIGN OF A BIOMACHINING PLANT FOR CONTINUOUS MICROETCHING
Diaz-Tena, Estibaliza, Rojo, Naiarab, Elías, Anaa, Aleksanyan, Aidaa,c, Gallastegui, Gorkab,
Barona, Astrida*
aDepartment of Chemical and Environmental Engineering. University of the Basque Country
UPV/EHU. Faculty of Engineering of Bilbao. c/ Alameda Urquijo s/n. 48013 Bilbao (Spain)
bDepartment of Chemical and Environmental Engineering. University of the Basque Country
UPV/EHU. University College of Engineering of Vitoria/Gasteiz. c/ Nieves Cano 12. 01006

Vitoria/Gasteiz (Spain)
cFaculty of Chemical Technologies and Environmental Engineering. National Polytechnic University of
Armenia. Teryan, 105, 0009 Yerevan (Republic of Armenia)

*Corresponding author: Email: astrid.barona@ehu.eus, Tel. 94 601 4110, Fax: 94 601 4179.
Biomachining is a microbiological process for selectively etching microstructures on a

workpiece by metal removal or dissolution. It has many advantages over the well-established
processes of chemical and physical machining, as microbial activity does not produce much
heat or residual stress on the piece, and metal removal is performed slowly and uniformly to
produce high quality and precision pieces. Additionally, energy consumption and capital and
labor requirements are low, and the hazardous compounds used in chemical processing are
replaced by microorganisms. Biomachining is therefore an affordable and sustainable
technology. Nevertheless, although biological processes are set to become part of future
industrial procedures for sustainable development, this process has yet to be implemented in
continuous production, and only lab or small-scale studies have been reported in the literature.
The aim here is to describe the overall design of a new biomachining plant for the continuous
production of etched microparts. The bioreactor consists of several consecutive adiabatic
vessels, with temperature, redox potential and pH control being ensured by remote-controlled
electrodes. A sulfuric acid solution for pH adjustment is fed automatically in response to a
signal from the pH electrode. Continuous mixing systems are assembled in all the vessels. An
electronic arm is programmed for immersing the parts in the active vessels, and biomass
regeneration periods (stand-by periods) are optimized. Finally, the operating range of the main
parameters and the detailed plans of the whole plant for continuous operation are presented.
Keywords: Biomachining; bioreactor; microbial etching; process design; control parameters
247
1st International Conference on Bioenergy & Climate Change
Towards a Sustainable Development
June 6th & 7th, 2016
http://bioenergy-climatechange.blogs.uva.es
Agrarian Engineering School

University of Valladolid – Soria Campus
www.ingenieriasoria.eu

Soria Preceedings

Uploaded by

Document Information

Copyright

Available Formats

Share this document

Share or Embed Document

Sharing Options

Did you find this document useful?

Is this content inappropriate?

Copyright:

Available Formats

Soria Preceedings

Uploaded by

Copyright:

Available Formats

See

The role of nitrogen availability on methane

Conference Paper · June 2016

Juan Carlos López Raul Muñoz

SEE PROFILE SEE PROFILE

Generation of gaseous biofuels from wastes View project

Algae biofuels and biorefinery View project

Available from: Juan Carlos López

The conference will focus on the latest developments on

Dr. Guillermo Quijano Dr. Alberto Ordaz

Organizing Committee Chairs

Fernando Fdz-Polanco – University of Valladolid (Spain)

Umberto Desideri – Università degli Studi di Pisa (Italy)

Luc Malhautier – École des mines d’Alès (France)

Carmen Gabaldón – University of Valencia (Spain)

Fédéric Thalasso – Cinvestav-IPN (Mexico)

Israel Díaz – University of Valladolid (Spain)

Evelyne Touraud – École des mines d’Alès (France)

Francisco Omil – University of Santiago de Compostela (Spain)

Flor Cuervo – UAM-Iztapalapa (Mexico)

Luis Miguel Romeo – University of Zaragoza & CIRCE (Spain)

Raúl Muñoz – University of Valladolid (Spain)

Francesco Fantozzi – Università degli Studi di Perugia (Italy)

Madalena Alves – University of Minho (Portugal)

Luis Hernández – University of Valladolid (Spain)

Alfons Stams – Wageningen University (Netherlands)

Alberto Ordaz – Mexiquense University (Mexico)

Pilar Lisbona – University of Valladolid (Spain)

Jorge Gómez – UAM-Iztapalapa (Mexico)

Daphne Hermosilla – University of Valladolid (Spain)

Pietro Bartocci – Università degli Studi di Perugia (Italy)

Pedro García – University of Valladolid (Spain)

Theo S.O. Souza – University of Sao Paulo (Brazil)

Guillermo Quijano – University of Valladolid (Spain)

Susana Calvo Ruiz – University of Valladolid

Adolfo Mercado Santamaría – University of Valladolid

José Ángel Miguel Romera – University of Valladolid

Luis Bonilla Morte – University of Valladolid

Ivonne Figueroa González – University of Valladolid

Daphne Hermosilla Redondo – University of Valladolid

Pilar Lisbona Martín – University of Valladolid

Epifanio Díez Delso – University of Valladolid

Israel Díaz Villalobos – University of Valladolid

Susana Gómez Ortega – Scientific Park, University of Valladolid

Alberto Ordaz Cortés (Chair) – Universidad Mexiquense del Bicentenario

Guillermo Quijano Govantes (Chair) – University of Valladolid

Speaker: Alfons Stams studies the anaerobic microbiology of environmental biotechnological

Oral Presentations – Bioenergy Production

A novelty approach in air pollution control:

2. MATERIALS AND METHODS

Figure 1: Conceptual diagram of anaerobic bioscrubber pilot system.

Table 1. Experimental plan for the removal of VOCs in a bioscrubber.

Specific surface, m-1 150 125

Water/air volume ratios 0.0035 - 0.0091 0.0076 - 0.0101

3. RESULTS AND DISCUSSION

3.1 Characterization of the foci

3.2 Plant Monitoring

Structural Insights on Ni-Ce-Zr Catalysts for

Manuel Benito1, Lucia Rodríguez2 and Gonzalo Muñoz2

Keywords: Biogas; Catalyst; Reforming; Hydrogen.

2. MATERIALS AND METHODS

2.1. Catalysts preparation

2.2. Catalysts characterization

Nitrogen adsorption-desorption isotherms at -196 ºC were obtained automatically by means