MTT is publishing its research findings in two series of publications: MTT Science and MTT Growth. The MTT Science series includes scientific presentations and abstracts from conferences arranged by MTT Agrifood Research Finland. Doctoral dissertations by MTT research scientists will also be published in this series. The topics range from agricultural and food research to environmental research in the field of agriculture. MTT, FI-31600 Jokioinen, Finland. email julkaisut@mtt.fi 29 Combination of biological and physico-chemical factors in the development of manure nutrient recovery and recycling-oriented technology Doctoral Thesis Anni Alitalo MTT CREATES VITALITY THROUGH SCIENCE www.mtt.fi/julkaisut Academic Dissertation: To be presented, with the permission of the Faculty of Agriculture and Forestry of the University of Helsinki, for public examination in lecture hall B2, Latokartanonkaari 7-9, Viikki on December 12th 2014, at 12 o’clock noon. Helsinki 2014 29 Combination of biological and physico-chemical factors in the development of manure nutrient recovery and recycling-oriented technology Doctoral Thesis Anni Alitalo Supervisors: Professor Erkki Aura Killintie 14, FI-31300 Tammela, Finland Professor Laura Alakukku University of Helsinki, Department of Agricultural Sciences/ Agrotehnology P.O.Box 28, FI-00014 Helsinki, Finland Reviewers Professor Sven G. Sommer University of Southern Denmark Institute of Chemical Engineering, Biotechnology and Environmental Technology Niels Bohrs Allé 1 DK-5230 Odense M, Denmark Professor Irini Angelidaki DTU ENVIRONMENT Department of Environmental Engineering Technical University of Denmark DK-2800 Kgs. Lyngby, Denmark Opponent Professor Willy Verstraete Ghent University Faculty of Bioscience Engineering Laboratory of Microbial Ecology and Technology Building A, room A0.092, Coupure Links 653, 9000 Ghent, Belgium ISBN 978-952-487-564-6 (Print) ISBN 978-952-487-565-3 (Electronic) ISSN 1798-1824 (Printed version) ISSN 1798-1840 (Electronic version) URN http://urn.fi/URN:ISBN:978-952-487-565-3 www.mtt.fi/mtttiede/pdf/mtttiede29.pdf Copyright MTT Agrifood Research Finland Anni Alitalo Distribution and sale MTT Agrifood Research Finland, Media and Information Services, FI-31600 Jokioinen, e-mail julkaisut@mtt.fi Printing year 2014 Cover image Risto T. Seppälä Printing house Tampereen Yliopistopaino Juvenes Print Oy MTT SCIENCE 29 3 Combination of biological and physico-chemical factors in the development of manure nutrient recovery and recycling-oriented technology Anni Alitalo MTT Agrifood Research Finland, Plant Production Research, FI-31600 Jokioinen anni.alitalo@mtt.fi Abstract Manure serves as an important source of nutrients, but is also an environmental concern. From the sustainable agriculture point of view it is of great importance to develop manure treatment technologies to recycle manure nutrients, and to reduce the odor and hygienic problems due to manure pro- duction and use. However, the develop- ment work has been greatly hampered by the complexity of manure as a raw materi- al and requires knowledge of various phys- icochemical and biological factors involved in its utilization. In this study, a sequential slurry manure processing scheme is described in which biological treatment was carried out in a series of continuously fed low-aerated tank reactors. Biological treatment was followed by ammonia stripping, conducted as a se- quential stripping procedure. The biolog- ical treatment served as a means of achiev- ing a pH increase of manure and made it possible to separate part of the nitrogen by stripping without chemical use. Limit- ed aeration during the biological treatment was applied in order to maintain nitrogen in ammonia form, thus preventing nitrate formation and excessive CO2 formation, and instead favoring humification process- es. Efficient odor reduction and good hy- gienic status were also among the target- ed objectives of the biological treatment. The functionality of the biological treat- ment was examined by using swine and dairy slurry manure. Experiments were carried out in pilot scale using six 600-L tanks connected in series with feedback. Before the actual start of the treatment, the treatment tanks were filled with mi- crobial seeding material. It was shown that the designed reactor system provid- ed stability for the process and increased treatment efficiency. However, efficient pretreatment was required in order to re- duce the dry matter content of the slur- ry manure to a level of 1−2% before the biological treatment. Part of the manure phosphorus was also removed with pre- treatment. The introduced treatment sys- tem reduced the manure odors to an un- detectable level or only very faint odor in four days. The pH-value at the end of the treatment increased above 8.5, and nitro- gen changes occurred mainly due to am- monia volatilization during the treatments. Nitrate formation was very low with the aeration rates used. Carbon reduction dur- ing the aeration treatment depended on the initial carbon content of the manure 4 MTT SCIENCE 29 and varied in the different studies between 11% and 57%. The treatment also caused changes in slurry precipitation characteris- tics, which were observed as a color change and a decline in the concentration of diva- lent cations and total phosphorus. It was also shown that the six tanks in series con- figuration with feedback served as a good device to achieve good hygiene of the end product. At best, over 90% reductions were observed in the numbers of enteric indicator organisms. However, the result obtained varied depending on the treat- ment run and indicator organism. It was shown that the buffer system in manure slurry was composed of TAN (to- tal ammoniacal nitrogen), CO2, HCO3− and CO32− and thus can be circumvented by making use of the ammonium-carbon- ate reduction obtained by biological treat- ment. Over 30% TAN removal by air stripping was shown to be possible without use of chemicals, if the pH of the biologi- cally treated swine manure was above 8.9. It can be concluded that this study ex- plains mechanisms and provides a basis for technologies used to efficiently reduce the manure odor, improve manure hygiene, and to separate nitrogen and phosphorus, enhancing the availability of manure frac- tions as fertilizer. The presented treatment method provides an option to increase re- cycling of the nutrients of manure and to reduce the odor and hygienic problems as- sociated with manure slurry. Keywords: Aeration, aerobic, ammonia, buffer, continuous, humification, hygiene, ni- trogen, odor, phosphorus, sequential, serial, slurry, stripping, treatment MTT SCIENCE 29 5 Biologisten ja fysikokemiallisten tekijöiden yhdistäminen lietelannan ravinteiden talteenottoon ja kierrätykseen tähtäävän teknologian kehittämisessä Tiivistelmä Lietelanta on tärkeä kasvinravintei-den lähde, mutta se on myös on-gelmallinen ympäristön kannal- ta. Maatalouden kestävyyden lisääminen edellyttää, että kehitetään lannan käsit- telymenetelmiä, jotka edesauttavat kier- rättämään lannan ravinteita ja vähentävät sen haju- ja hygieniaongelmia. Menetel- mäkehitystä on kuitenkin huomattavasti vaikeuttanut lannan kompleksisuus raaka- aineena, mikä edellyttää fysikaalis-kemial- listen ja biologisten tekijöiden ymmärrystä ja yhteensovittamista käsittelymenetelmiä kehitettäessä. Tässä tutkimuksessa kehitettiin vaiheit- tainen lietteen käsittelymenetelmä, jossa biologinen käsittely toteutettiin rajoitetus- ti ilmastetuissa, toisiinsa sarjaan kytket- tyjen reaktorien järjestelmässä. Biologis- ta käsittelyä seurasi ammoniakin erotus ilmavirran avulla täytekappalekolonnis- sa strippaamalla. Erotusta tehostettiin toistamalla strippauskertoja. Lietelannan pH-arvo nousi biologisen käsittelyn ai- kana, mikä mahdollisti osittaisen typen erotuksen strippaamalla ilman kemikaa- leja. Ilmastus pidettiin biologisen käsitte- lyn aikana maltillisena, jotta typpi säilyisi ammonium-muodossa ja nitraatin muo- dostuminen olisi mahdollisimman vähäis- tä. Lisäksi ilmastusta rajoittamalla pyrit- tiin hillitsemään hiilidioksidipäästöjä ja suosimaan humifioitumisreaktioita. Ta- voitteena oli myös tehokas hajun vähentä- minen ja lietelannan hygieenisyyden pa- rantaminen biologisen käsittelyn aikana. Biologista käsittelyä tutkittiin sian ja nau- dan lietelannoilla. Tutkimuksessa käytet- ty biologinen käsittelyjärjestelmä koostui kuudesta sarjaan kytketystä jatkuvasyöt- teisestä ilmastetusta prosessitankista, jois- ta viimeisen ja ensimmäisen välillä oli takaisinkytkentä. Kokeet toteutettiin pi- lot-mittakaavassa 600 litran tankeissa. En- nen varsinaista käsittelyä, prosessisäiliöt oli täytetty etukäteen tuotetulla mikrobiym- pillä. Kehitetty reaktorijärjestelmä osoit- tautui toiminnaltaan vakaaksi ja tehok- kaaksi. Biologisen käsittelyn edellytys oli esikäsittely, jolla lietteen kuiva-ainepitoi- suus laskettiin 1-2 %:n tasolle. Neljän päi- vän prosessoinnin jälkeen biologinen kä- sittely joko poisti lietteen hajun täysin tai haju oli vain hyvin heikosti havaittavissa. Käsittelyn lopussa lietteen pH-arvo nousi 8,5 yläpuolelle ja typen muutokset tapah- tuivat lähinnä ammoniakin haihtumise- na käsittelyjen aikana. Nitraattia ei juuri muodostunut käytetyillä ilmastusmääril- lä. Hiilen häviö ilmastuskäsittelyn aikana Anni Alitalo MTT, Kasvintuotannon tutkimus, 31600 Jokioinen anni.alitalo@mtt.fi 6 MTT SCIENCE 29 riippui alkuperäisestä lietteen hiilipitoisuu- desta ja vaihteli eri kokeissa 11 ja 57 pro- sentin välillä. Käsittely muutti myös liet- teen saostusominaisuuksia. Ne havaittiin lietteen värimuutoksena ja sen kaksiarvois- ten kationien ja kokonaisfosforin pitoisuu- den laskuna. Tulosten mukaan biologinen käsittely paransi lietteen hygienistä laa- tua. Parhaimmillaan käsittely vähensi yli 90 % suolistoperäisten indikaatiomikrobi- en määrää. Saavutettu tulos vaihteli kui- tenkin sen mukaan, miten lanta käsiteltiin ja mitä indikaatio-organismia käytettiin. Tutkimuksessa osoitettiin, että lietteen puskurisysteemi koostui kokonaisammo- nium-ammoniakkitypestä, hiilidioksidis- ta, bikarbonaatista ja karbonaatista, ja että se on kierrettävissä hyödyntämällä biolo- gisen käsittelyn aikana saavutettua am- monium-bikarbonaattisysteemin pusku- rikyvyn vähenemistä. Tulosten mukaan lietteen kokonaisammoniumtypestä voi- tiin poistaa yli 30 % ilman kemikaalikä- sittelya strippaamalla, kun biologisesti kä- sitellyn lietteen pH-arvo oli yli 8,9. Yhteenvetona voidaan todeta, että tämä tutkimus selittää edellä esitettyjä mekanis- meja ja luo perustan teknologioille, joilla voidaan tehokkaasti vähentää lietelannan hajua, parantaa sen hygieniaa ja lisätä liet- teen jakeiden hyödynnettävyyttä lannoit- teina erottamalla typpi ja fosfori. Esitetyt käsittelymenetelmät luovat mahdollisuu- den parantaa lannan ravinteiden kierrätys- tä ja vähentää lietelantaan liittyviä haju- ja hygieniaongelmia. Avainsanat: Aerobinen, ammoniakki, fosfori, haju, hygienia, humifioituminen, ilmas- tus, jatkuva, käsittely, lietelanta, pe- räkkäinen, puskuri, sarja, strippaus, typpi MTT SCIENCE 29 7 Acknowledgements The early steps towards this thesis were taken already in 2005 when I joined the Slurry manure odor removal and frac- tionation project led by Professor Erkki Aura at the Soils and Environment unit of MTT Agrifood Research Finland. The three-year project was funded by the Finn- ish Ministry of Agriculture and Forest- ry. During this time period early stage pi- lot scale experiments with swine slurry as well as the ammonia stripping experiments were conducted. Pilot scale experiments with dairy slurry were continued thereaf- ter with the funding of Marjatta ja Eino Kollin Säätiö. Several people were involved in the exper- imental and laboratory work of the exper- iments. I wish to thank them all for their excellent work. I am especially grateful to Senior Research Technician Risto T. Sep- pälä, who has been the person behind all the technical construction/equipment and has been responsible for running of con- tinous processes. I would like to express my gratitude to Mrs. Katariina Saarela, who has been responsible for the laborato- ry analyses and running of the continous processes. I am grateful to my coauthors Tuomas Pelto-Huikko, Aleksis Kyrö and Johanna Nikama for carrying out most of the experiments and to Doctor Tapio Salo who helped me with statistical methods. I also wish to thank D.Ph. Minna Kahala and Laboratory Engineer Anneli Virta for their advice and assistance with biotech- nological methods. There are no words that can express my gratitude to Professor Erkki Aura for his ever inspiring guidance, pulping new ideas and deep scientific understanding. I sincerely thank Professor Laura Alakuk- ku, who has supported me since the be- ginning troughout this project and offered valuable proposals for improvements on this thesis and contributed greatly to shap- ing the work into its final form. I grateful- ly acknowledge Professor Irini Angelidagi and Professor Sven G. Sommer for their insightful pre-examination of the manu- script. I also thank Professor Martti Esala and Director Lic.Sc.Agric. Markku Järv- enpää for their supporting attitude towards this work. This study was closely related to a larg- er R & D project, aimed at the develop- ment and market introduction of a new kind of slurry manure treatment system. I am grateful to Mr. Juha Takala for pro- viding inspiring and broad vision into this subject area. The multidisciplinary approach pre- sented in this thesis would not have been possible without long term fund- ing granted by Teknillis-yhteiskunnallin- en tutkimussäätiö. I am also grateful to other financers: The Ministry of Agricul- ture and Forestry, Marjatta ja Eino Kollin Säätiö and the Academy of Finland (con- tract No. 130803). Finally, I wish to thank my family for their love and enduring patience during the work associated with this thesis. 8 MTT SCIENCE 29 List of original publications This thesis is based on the following publications: I Alitalo, A., Pelto-Huikko, T. & Aura, E. 2013. Application of a Series of Continu- ously Fed Aerated Tank Reactors System for Recycling of Swine Slurry Nutrients. Journal of Sustainable Development 6: 26-38. II Alitalo, A., Kyro, A. & Aura, E. 2012. Ammonia stripping of biologically treated liquid manure. Journal of Environmental Quality 41: 273–280. III Alitalo, A., Alakukku, L. & Aura, E. 2013. Process design and dynamics of a series of continuously fed aerated tank reactors treating dairy manure. Bioresource Tech- nology 144: 350-359. IV Alitalo, A., Nikama, J. & Aura, E. 2014. Fate of faecal indicator organisms and bac- terial diversity dynamics in a series of continuously fed aerated tank reactors treat- ing dairy manure. Ecological Engineering. Submitted. The publications are referred to in the text by their roman numerals. MTT SCIENCE 29 9 The following table presents the contributions of the authors to the original articles of the dissertation: I II III IV Initial idea EA AA, EA AA AA, EA Planning the experiment TP, EA AA, EA, AK AA, EA AA, JN Conducting the experiment TP, AA AK AA JN Data analysis AA AA AA AA, JN Manuscript preparation AA, EA AA, EA AA, EA, LA AA, EA AA = Anni Alitalo EA = Erkki Aura TP = Tuomas Pelto-Huikko AK = Aleksis Kyrö JN = Johanna Nikama LA = Laura Alakukku Contributions 10 MTT SCIENCE 29 Abbreviations DM dry matter CMBRs completely mixed batch reactors CMFRs completely mixed flow reactors CMR complete mix reactor COD chemical oxygen demand DO dissolved oxygen ED electrodialysis HRT hydraulic retention time IC inorganic carbon MF microfiltration NF nanofiltration Nsol soluble N Ntot total N ORP oxidation-reduction potential PFRs plug flow reactors Pi inorganic phosphorus Po organic phosphorus Ptot total phosphorus RO reverses osmosis TAN total ammoniacal nitrogen (NH3 and NH4+) TC total carbon TOC total organic carbon TS total solids UF ultrafiltration VFA volatile fatty acids VS volatile solids Contents 1 Introduction ............................................................................................................ 13 1.1 Characterization of animal manure ..........................................................................13 1.1.1 Concentrations of plant nutrients ..................................................................14 1.1.2 Characteristic components of manure ............................................................15 1.1.3 Manure containing P compounds .................................................................16 1.1.4 Odor compounds ..........................................................................................17 1.1.5 Buffer capacity ...............................................................................................18 1.2 Manure volumes, environmental impacts and legislation .........................................19 1.2.1 Manure volumes and regional concentration .................................................19 1.2.2 Environmental impacts related to animal manure ...........................................21 1.2.3 Environmental legislation ...............................................................................25 1.3 Manure treatment technologies ...............................................................................25 1.3.1 Method categorization into physical, chemical and biological processes ..........26 1.3.2 Phosphorus removal and recovery technologies ..............................................26 1.3.3 Treatment of the liquid fraction after solid-liquid separation ..........................28 1.3.4 Odor treatment technologies ..........................................................................30 1.3.5 Treatment methods for inactivation of manure pathogens .............................31 1.4 Needs for manure treatment and the importance of nutrient recycling .....................33 1.5 Theoretical background of the Manure treatment system of the present study ..........34 1.5.1 Reactor types .................................................................................................34 1.5.2 Degradation and recondensation processes involved in biological treatment ..35 1.5.3 Physico-chemical factors involved in treatment processes ..............................37 2 Objectives of the study ........................................................................................... 41 3 Materials and methods ............................................................................................ 43 3.1 Properties of raw slurries ..........................................................................................43 3.2 Technical description of the treatment system and treatment processes ...................43 3.2.1 Pre-treatment processes ..................................................................................43 3.2.2 Serial treatment system ..................................................................................44 3.3 Experimental designs ..............................................................................................45 3.4 Analytics .................................................................................................................45 3.4.1 Chemical methods and buffer capacity ...........................................................45 3.4.2 Microbiological methods ................................................................................48 3.4.3 Data analysis and statistics ..............................................................................48 4 Results .................................................................................................................... 50 4.1 Effects of pre-treatment on manure composition ....................................................50 4.2 Biological treatment ................................................................................................50 4.2.1 Operational conditions, dynamics and stability of the serial treatment system 50 4.2.2 Physico-chemical changes due to treatment ....................................................53 4.2.3 Effect of treatment on manure nutrient content .............................................55 4.2.4 Total bacterial counts and enteric indicator organisms in slurry .....................55 4.2.5 Bacterial diversity dynamics ...........................................................................57 4.3 Buffer system in the slurry and ammonia stripping .................................................58 4.3.1 Ammonia stripping .......................................................................................58 4.3.2 Buffer system composition in the slurry ........................................................60 12 MTT SCIENCE 29 5 Discussion ............................................................................................................... 61 5.1 Solids separation before aeration treatment ..............................................................61 5.2 Theoretical basis for the designed reactor type .........................................................62 5.2.1 Several bioreactors grouped by serial connection ............................................62 5.2.2 Feedback ........................................................................................................62 5.2.3 Use of limited aeration ...................................................................................62 5.3 Benefits of the serial treatment system .....................................................................63 5.3.1 Treatment stability towards system failures .....................................................63 5.3.2 Treatment efficiency .......................................................................................63 5.3.3 Odor reduction ..............................................................................................64 5.3.4 Advanced process control ...............................................................................64 5.4 Process conditions – Biological and physico-chemical factors in a serial system .......65 5.4.1 Treatment start-up and inoculation ................................................................65 5.4.2 Aerobic decomposition under limited aeration and humification ..................67 5.4.3 pH and its relationship with the release of NH3 and CO2 from the system ...69 5.5 The role of manure buffer system in manure treatment ............................................69 5.6 Sequential nitrogen separation .................................................................................70 5.7 Hygiene ...................................................................................................................71 5.8 Applicability of the developed manure treatment technology ...................................72 6 Conclusions ............................................................................................................ 75 7 References ............................................................................................................... 77 MTT SCIENCE 29 13 1 Introduction Slurry manure is a by-product of food production, composed mainly of a mixture of feces and urine of live- stock animals. During the course of his- tory its status has changed from the origi- nally valuable plant nutrient to disposable waste and then once again become appre- ciated as a raw material, the value added components and energy content of which should be recovered and recycled in the most effective way (Sommer et al., 2013). Changes in the approach to slurry revolve around global key issues: population size increase, globalization, environmental emissions and climate change, limited nu- trient resources, and depleted fossil fuel re- serves (Gilbert, 2009; Fedoroff et al., 2010; Godfray et al., 2010; Sutton et al., 2011; Tian et al., 2012). Global population size has been estimat- ed to increase from approximately seven billion today to probably over nine bil- lion by 2050, and to level off somewhere between 9 and 12 billion people by the end of the century (UN, 2012). This will be reflected in an increase in demand for food production (Godfray et al., 2010). At the same time, the increasing standard of living is further forecasted to increase this demand, although the production of a varied, high-quality diet including ani- mal protein is known to require additional resources (FAO, 2009). Crops may also be used for biofuels and industrial purposes (Godfray et al., 2010). It is likely that com- petition for land, water and energy will in- tensify while the effects of climate change will become increasingly evident (Tilman et al., 2001; WRI, 2005). Limited natural nutrient resources and pressing environmental issues increase the challenges facing food production. Al- though phosphorus (P) resources are rel- atively abundant and significant globally, experts disagree on how much phosphate is left and how quickly it will be exhaust- ed (Gilbert, 2009). In the case of nitrogen (N), industrial nitrogen fixing utilizes fos- sil fuels (Smil, 2001), which are a dimin- ishing resource. Limited resources increase fertilizer prices. On the other hand, waste- ful use of resources causes a global en- vironmental threat both to water bodies and to the atmosphere (Sutton et al., 2011; Tian et al., 2012). It is obvious that the growing need for food requires more intensive food produc- tion, which should be carried out in an ecologically, environmentaly, economi- cally, and socially sustainable manner. In the future, manure should be managed more intelligently than at present, with a focus on high energy recovery, increas- ing recycling of plant nutrients, and using the most recalcitrant organic matter for soil carbon sequestration, thereby limit- ing negative impacts on soil, air and water quality (Jensen et al., 2013) without for- getting hygiene aspects and nuisance odors (e.g. Arnold et al., 2006; Bicudo and Go- yal, 2003; Newell et al., 2010; Schiffman et al., 2004). 1.1 Characterization of animal manure Manure is composed not only of the urine and feces of livestock, but also of the bed- ding, spilled feed, water used for washing, and other materials mixed with it. Live- stock manure is a complex mixture from a chemical, physical and biological point of view. The content is largely dependent 14 MTT SCIENCE 29 on the digestion of the specific feed by the individual animal (Carter and Hae- Jin, 2013). During storage, the manure characteristics change continuously due to further degradation (e.g. Zhu et al., 2001; Møller et al., 2002; Popovic and Jensen, 2012). All these factors may influence the manure properties resulting from differ- ent treatments. Manure dry matter content is an impor- tant physical property affecting e.g. the viscosity and technical handling of ma- nure (MWPS, 2004; Sommer et al., 2013). Manure has been classified according to its dry matter content as liquid, slurry, (semi-solid) and solid manure, but defi- nitions have differed between countries (Kemppainen, 1989; MWPS, 2004; Som- mer et al., 2013): • Liquid manure contains a very low dry matter content and can be handled with irrigation equipment. • Slurry manure contains 3−10% dry mat- ter content. It flows under gravity and can be pumped, but may require special pumps for handling. Semi-solid manure contains between 10 and 20% solids. Semi-solid manure is too thick to pump. • Solid manure contains 18−25% dry mat- ter content or more. Important raw material properties of ma- nure are nutrient and other constituent concentrations. When manure handling technology is developed, the characteriza- tion of particle size distribution, nutrient fractions, odor combounds, and manure buffer capacity is often relevant. 1.1.1 Concentrations of plant nutrients Manure has usually been characterized from the plant nutrient perspective by analyzing the basic elements of interest, mainly the physical properties of the dry matter content, pH, and the contents of the nutrients N, P and potassium (K). Additionally, calcium (Ca), magnesium (Mg) and micronutrients such as sodi- um (Na), chlorine (Cl), iron (Fe), copper (Cu), zinc (Zn), manganese (Mn) and bo- ron (B) have been analyzed (Kemppainen, 1989). A comprehensive study of the ma- nure nutrient content in Finland was presented by Erkki Kemppainen in his doctoral thesis (Kemppainen, 1989). In addition, a Finnish commercial compa- ny, Eurofins Viljavuuspalvelu Oy, collects statistics on the manure nutrient content of the samples sent by farmers for fertili- ty analyses (Viljavuuspalvelu, 2014). Slur- ry manure nutrient contents in the late 1980s according to Kemppainen (1989) and some twenty years later (2006–2009) are presented I Tables 1 and 2, based on Table 1. Average pH and contents of dry matter (DM), total N (Ntot), soluble N (Nsol), total phosphorus (Ptot) and potassium (K) in slurry manure according to Kemppainen (1989). Manure type DM pH Ntot Nsol Ptot K (%) (g/kg (fresh weigth)) cow slurry 8.1 7.0 3.3 1.8 1.0 2.8 pig slurry 9.2 7.0 5.4 3.6 1.9 2.0 Table 2. Average pH and contents of dry matter (DM), total N (Ntot), soluble N (Nsol), total phosphorus (Ptot), potassium (K), magnesium (Mg), calcium (Ca), and sodium (Na) in slurry manure according to Vil- javuuspalvelu (2014). Manure type DM Ntot Nsol Ptot K Mg Ca Na (%) (g/kg (fresh weigth)) cow slurry 5.5 3.0 1.7 0.5 2.9 0.5 0.8 0.3 pig slurry 3.5 3.7 2.4 0.8 1.8 0.5 1.1 0.5 MTT SCIENCE 29 15 the Viljavuuspalvelu statistics (Viljavuus- palvelu, 2014). By comparing these data, a significant difference was observed especially in pig slurry dry matter (DM) and phosphorus contents (Tables 1 and 2). In part, this was probably due to the differences in sampling techniques between these data. The samples of Viljavuuspalvelu statis- tics represented a larger number of sam- ples taken by the farmers themselves, and therefore the data was possibly more het- erogenous. Secondly, the differences be- tween the two data sets were probably caused by the differences in farming prac- tices and animal diets in the late 1980s compared to the 2000s. 1.1.2 Characteristic components of manure Fiber and other constituent compositions Manure processing, biogasification, and the interest in separate value-added com- ponents from manure have increased the need for detailed chemical information concerning animal manures (Chen et al., 2003). The fiber composition, including the content of cellulose, hemicellulose, and lignin in manure, can be determined by Van Soest’s fiber analysis test using a reflux apparatus (Goering and Van Soest, 1970). In addition to the fiber composition, in some studies manure elemental composi- tion, and also protein and even amino acid content have been characterized (e.g. Chen et al., 2003). In cattle manures fiber was the main com- ponent, as the total lignocellulosic mate- rials content in dairy manure was more than half of the dry matter (Table 3). Pop- ovic (2012) concluded that slurry parti- cles consist of 10–15% proteins, 15–20% humic substances/lignin, 10–30% carbo- hydrates/cellulose, 10–25% hemicellulose, 5% fat and 15–25% inorganic compounds such as struvite based on the data obtained from Liao et al. (2004), Rico et al. (2007); and Christensen et al. (2009). Fiber pro- vided the substantial resources of cellulose and hemicellulose, being degradable into monosaccharides for use as feedstocks to produce value-added products such as gly- cols and diols (Liao et al., 2004). Particle size distribution Different compounds of manure fall into different particle sizes affecting, for exam- ple, the composition of fractions in sepa- ration. Particles of diameter <10 µm con- tained alkylaromatics, phenols and lignin monomers, carbohydrates, and N-con- taining compounds (e.g. peptides), where- as larger particles (10 µm – 2 mm) most- ly contained fatty acids, lignin dimers and sterols in pig slurry (Aust et al., 2009). Ac- cording to Masse et al. (2005), particles smaller than 10 µm represented 64% of DM in raw swine manure, and 50% of to- tal P was associated with particles between 0.45 and 10 µm. Only 30% of the P was linked to particles larger than 10 µm and approximately 95% of organic N was as- sociated with particles between 0.45 and Table 3. Comparison of mean fiber and protein contents (%, DM) in cattle and swine manures accord- ing to Chen et al. (2003). Manure type Crude protein Total fiber Hemicellulose Cellulose Lignin (%, DM) Cattle dairy 18.1 52.6 12.2 27.4 13.0 Swine nursery 25.1 39.2 21.9 13.2 4.1 Swine grower 22.7 40.8 20.5 13.9 6.4 Swine finisher 22.0 39.1 20.4 13.3 5.4 16 MTT SCIENCE 29 10 µm (Masse et al., 2005). These values vary in the literature (Masse et al., 2005; Popovic, 2012) due to differences in ani- mal feeding practices, management, and slurry storage practices. 1.1.3 Manure containing P compounds In most studies of manure P composi- tion, inorganic and total P have been de- termined. In order to improve the man- agement of manure P, new methods have been elaborated to identify and quantify manure organic P forms (e.g. He and Hon- eycutt, 2001; Turner and Leytem, 2004). Hedley fractionation The sequential extraction schemes origi- nally developed for soil phosphorus char- acterization have also been introduced to fractionate manure phosphorus (e.g. Yli- vainio et al., 2008). The widely used Hed- ley fractionation procedure (Hedley et al., 1982) aims to categorize P into pools based on biological availability. The inorganic- P (Pi) and organic-P (Po) extracted with stronger solutions are assumed to repre- sent less bioavailable P pools than the pre- ceding fractions extracted with milder so- lutions (Table 4). The first step in Hedley fractionation extracts loosely bound water- soluble P. The second extraction is made with sodium bicarbonate (NaHCO3); this extraction has been suggested to provide an estimate of plant available P. These first two fractions constitute the labile and readily plant available fraction of P, and the next two extractions, made with so- dium hydroxide (NaOH) and hydrochlo- ric acid (HCl), represent P pools which are bound more strongly. The remaining frac- tion is the residual fraction (Hedley et al., 1982; Sharpley and Moyer, 2000). Ylivain- io et al. (2008) reported that in dairy ma- nure, 81% of the sum of the P fractions was water-soluble (Table 4). A notable pro- portion (14%) of the P present was organic P and most of it was water- and NaHCO3- extractable, totalling about 12% of the to- tal P fractions. After sequential fractionation, organic phosphorus fractions have been fractionat- ed into even more specific organic P forms with orthophosphate-releasing enzymes. According to the study of He and Honey- cutt (2001), pig and cattle manures were first sequentially fractionated into water- soluble P, NaHCO3-soluble P, NaOH-sol- uble P, HCl-soluble P, and residual P. Part of the organic P in these fractions could be identified by the enzymatic treatments as phytate (39% for pig manure and 17% for cattle manure in water-soluble organ- ic P), simple phosphomonoesters (43% for pig manure and 15% for cattle manure in NaOH-soluble organic P), nucleotide-like phosphodiesters (2–12%), and nucleotide pyrophosphate (0–4%). NMR spectroscopy Nuclear magnetic resonance (NMR) spec- troscopy can provide compound-specif- ic information on manure phosphorus. Both solid-state and solution 31P NMR Table 4. Concentrations of inorganic and organic P sources according to the Hedley fractionation scheme in air dried dairy manure (Ylivainio et al., 2008). Extractant Inorganic-P (Pi) (mg/g) Organic-P (Po) (mg/g) Water 3.2 0.3 NaHCO3 0.2 0.2 NaOH 0.1 0.1 HCl 0.2 Σ 3.7 0.6 MTT SCIENCE 29 17 spectroscopy have been used. Turner and Leytem (2004) used a two-step extrac- tion procedure and NMR spectroscopy to quantify phosphorus compounds in ex- tracts of swine and cattle manure. Initial extraction in NaHCO3 recovered readi- ly soluble phosphorus, whereas a second extraction in NaOH-EDTA recovered poorly soluble phosphorus. Organic phos- phorus in the readily soluble fraction in- cluded DNA, phospholipids, and simple phosphate monoesters in both manures, whereas the poorly soluble fraction includ- ed poorly soluble phosphate, plus phytic acid in swine manure and a range of phos- phate monoesters and diesters in cattle ma- nure (Turner and Leytem, 2004). 1.1.4 Odor compounds Manure odors are a complex mixture of volatile fatty acids (VFA), alcohols, ar- omatic compounds, amides (including NH3 ), and sulfides (Hartung and Philips, 1994). In a study of Schiffman et al. (2001, Table 5) a total of 411 compounds were found in odorous emissions from swine facilities. Odorous gases and volatile compounds are produced during incomplete anaerobic fer- mentation (Mackie et al., 1998). Anaerobic microorganisms use organic compounds as their electron donor and as sources for cell synthesis and metabolism (for energy and carbon), during which various odor- Table 5. Heterogenity of odorous compounds. Compound group, example compound of the group, chem- ical formula, and odor characteristic of the compound (modified from Shiffman et al., 2001). Compound group Compound Formula Odor characteristics Acids Formic acid Acetic acid Propionic acid HCOOH CH3COOH CH3CH2COOH Irritant, purgent Irritant, purgent Irritant, purgent Alcohols Methanol Ethanol CH3OH Alcoholic Aldehydes Formaldehyde Acetaldehyde Benzaldehyde HCHO CH3CHO C6H5CHO Pungent, rotten Pungent Almond. irritant Amides Acetamide N,N-dimethylformamide CH3CONH2 HCON(CH3)2 Irritant, fishy, pungent Amines Methylamine (aminomethane) CH3NH2 Initant, putrid, fishy Aromatics Benzene Toluene Methylstyrene C6H6 C6H5CH3 C6H4(CH3)CH=CH2 Benzene-like Irritant Esters Mehyl formate HCOOCH3 Irritant Ethers Diethyl ether Furan C2H5OC2H5 C4H4O Sweet, pungent, irritant Fixed gases Ammonia NH3 Sharp, pungent Halogenated hydrocarbons Chloroform CHCl3 Hydrocarbons 2-methylbutane CH3CH2CH(CH3)2 Irritant Ketones 2-propanone CH3COCH3 Irritant Nitriles Benzen acetonitrile Aromatic Other N containing compounds Pyridine C5H5N Irritant, burnt Phenols Phenol C6H5OH Irritant Sulfur containing compounds Hydrogen sulfide H2S Rotten eggs 18 MTT SCIENCE 29 ous gases and volatile compounds are pro- duced (e.g. Hartung and Philips, 1994). The quantity and variety of the organic matter contributes directly to odor gen- eration and is linked to animal type, diet, manure composition and microbial fer- mentation (Miller and Varel, 2003). Starch and protein are primary substrates for odor compounds produced. Starch fermentation has been shown to dominate in cattle ma- nure fermentation, whereas both protein and starch fermentation occurred in swine manure (Miller and Varel, 2003). More of- fensive compounds (branched chain vol- atile fatty acids and aromatic ring com- pounds) tended to be produced during protein fermentation, emphasizing the sig- nificant malodor related to swine manure (Miller and Varel, 2003). Volatilized fatty acids and aromatic com- pounds have been most closely correlat- ed to odor (Zhu et al., 1997; Powers et al., 1999; Zahn et al., 2001), and the use of VFA level to determine the odor inten- sity of swine manure is commonly uti- lized (Evans et al., 1986; Miller and Varel, 2001). Another commonly used method to measure odor is olfactometry, by which the odor concentration is measured as di- lution-to-threshold in an olfactometer with human panelists (CEN, 2003). Recently, also direct on-site measurements of odor and chemical measurements by proton- transfer-reaction mass spectrometry (PTR- MS) have been used (e.g. Hansen et al., 2012). 1.1.5 Buffer capacity Manure buffer capacity has an essential role in manure treatment processes. This is due to the buffer system resistance to pH changes, which complicate solids and P removal processes as well as N separa- tion (Paul and Beauchamp, 1989; Husted et al., 1991; Sommer and Husted, 1995). High amounts of chemicals are required, which often make the treatment econom- ically unprofitable. Buffers are compounds that resist changes in pH upon the addition of acids or bas- es. Buffer systems are usually composed of a weak acid or base and its conjugate salt (Holman et al., 2013). The components act in such a way that addition of an acid or base results in the formulation of a salt, causing only a minor change in pH (Hol- man et al., 2013). Buffer solutions achieve their resistance to pH change because of the presence of an equilibrium between the acid HA and its conjugate base A-.HAH+ + A- (1) When strong acid is added to an equi- librium mixture of the weak acid and its conjugate base, the equilibrium is shifted to the left, in accordance with Le Chat- elier’s principle (e.g. Holman et al., 2013). Because of this, the hydrogen ion (H+) concentration increases by less than the amount expected for the quantity of strong acid added. Similarly, if strong alkali is added to the mixture the hydrogen ion concentration decreases by less than the amount expected for the quantity of alka- li added (Holman et al., 2013). The complex chemical buffer system in manure consists of a number of acids and bases, the most important of which are total ammoniacal nitrogen (TAN), total inorganic carbon (CO2, HCO3−, CO32−), volatile fatty acids, and other organic compounds (Paul and Beauchamp, 1989; Husted et al., 1991; Sommer and Hus- ted, 1995). The major buffer reactions in manure are (Husted et al., 1991; Sommer and Hus- ted, 1995): Enzymatic hydrolysis of urea originat- ing from animal urine to ammonium carbonate: CO(NH2)2 Urease, H2O  2 NH4+ + CO32- (2) MTT SCIENCE 29 19 Decomposition of the ammonium carbon- ate into NH3 and CO2 gases: 2NH4+ + CO32-  NH4+ + HCO3- + NH3 (3) NH4+ + HCO3-  CO2 + H2O + NH3 (4) If the released gases are removed, equa- tions (3) and (4) will proceed to the right. The equations show the combined effect of NH4+ hydrolysis generating H+, and the release of CO2, which in turn requires H+; on the one hand protons are released in the hydrolysis of the ammonium, and on the other hand the liberation of carbon diox- ide binds protons: Ammonia hydrolysis: NH4+  NH3 + H+ (5) Release of carbon dioxide: CO32- + H+  HCO3- (6) HCO3- + H+  CO2 + H2O (7) Because of the low equilibrium constant of equation (5) (pKaNH4 = 9.3; Stumm and Morgan, 1981), volatilization of ammoni- acal N only occurs under neutral to alka- line conditions. Being itself an acidifying process, NH3 volatilization is enhanced by a high H+ buffer capacity (Avnimelech and Laher, 1977; Vlek and Stumpe, 1978). Balancing H+ release and uptake, the sim- ple (NH4)2CO3 system described in equa- tions (2) to (7) allows complete simultane- ous volatilization of NH3 and CO2. VFAs may volatilize or be removed by aerobic microbial decomposition, there- by consuming H+ (Paul and Beauchamp, 1989): CH3COO- + H+  CH3COOH (8) 1.2 Manure volumes, environmental impacts and legislation 1.2.1 Manure volumes and regional concentration Agricultural intensification, regional spe- cialization, concentration of production on a decreasing number of farms, and in- creasing farm size during recent decades have been intensive particularly in Fin- land (Niemi and Ahlstedt, 2010), but also in other parts of Europe too. In Finland, the regionally differentiated agricultural support has affected the specialization of agriculture and has in particular led to re- gional concentration of livestock produc- tion (VTV, 175/2008). Number of livestock farms and animals In 2012, there were 1712 pig farms in Fin- land with 1.3 million pigs. The number of cattle farms was 13321 with 912,800 animals, of which 31% were dairy cows (Tike, 2013). Pig farms were mainly lo- cated in south-western Finland (27%), in southern Ostrobothnia (17%) and in Os- trobothnia (16%), whereas the majority of the dairy farms were located in northern Ostrobothnia (15%) and in northern Savo (13%). The number of pig farms decreased by 72% between 1995 and 2012 but the number of pigs remained almost the same (-8%, Tike, 2013). During the same time period, the number of dairy farms de- creased by 70% and the number of dairy cattle by 29%, but milk production re- mained on approximately the same level (Niemi and Ahlsted, 2013; Tike, 2013). On the European level, pig production has been concentrated in a few countries, with Denmark, Germany, Spain, France, the Netherlands and Poland having more than two thirds of the breeding pigs be- tween them (Fig. 1, European commis- sion, 2010). In 2011, the EU-27 had 86.2 million bo- vine animals. About third of them (22.8 20 MTT SCIENCE 29 million) were dairy cows (representing 27% of the total bovine population of the EU-27). The majority of the dairy cows were located in three different countries, i.e. in Germany (18%), France (16%) and the United Kindom (>8%) (European Commission, 2013). Manure volumes and storage In Finland, the annual amount of ma- nure generated is not monitored statis- tically. However, based on animal num- bers and official manure storage capacities, manure production was estimated to be 13  543  967 tons/year (fresh weight) in 2009 (Grönroos et al., 2009; Luostarinen and Grönroos, 2013). In terms of manure production, the most important animals were cattle and pigs, which produced more than 95% of the total amount of manure generated. Of the total amount produced in 2009, 39% was cattle slurry, 45% cattle solid, 11% pig slurry and 4% pig solid ma- nure (Luostarinen and Grönroos, 2013). The entire manure production in the EU that is potentially available for manure processing has been estimated to be 1.4 billion ton/year (Foged et al., 2011). The highest production has been in France, fol- lowed by Germany, and the lowest in Mal- ta (Foged et al., 2011). In Finland, the manure storage must be sized according to the annually generated amount of manure except for pasture ma- nure (VNa 931/2000). With increasing production levels, manure outputs have also increased. For example, for dairy cows the increase was estimated to be 44% from Source: Eurostat (agr_r_animal) Figure 1. Number of sows by region in 2008. 1 dot = 1,000 sows – NUTS 2 except DK, DE, UK (NUTS 1) (European commission, 2010). MTT SCIENCE 29 21 1990 to 2012 (Helstedt et al., 2013, Table 6). According to Hellstedt et al. (2013), the water volume used for e.g. washing was 9% of the total amount of slurry in the case of dairy cows and sows. 1.2.2 Environmental impacts related to animal manure Agriculture has been identified as a signif- icant contributor to nutrient losses to the environment, especially from livestock ma- nure. Globally, livestock has been evaluated to excrete about 100 Tg N per year, but only 20–40% of this amount is estimated to be recovered and applied to crops (Sheldrick et al., 2003; Oenema and Tamminga, 2005). The remainder was dissipated into the envi- ronment. The estimated amounts of phos- phorus (P) and potassium (K) in livestock manure were 1.5 and 3 times the amounts of P and K in mineral fertilizers, but only a fraction of manure P and K was efficiently utilized (Sheldrick et al., 2003). Livestock production systems exert various influences on the environment, distribut- ing emissions into the air, soils and water- courses (Fig. 2). The influences on the en- vironment greatly depend on the livestock production system itself, the management, and the environmental conditions (Oene- ma et al., 2007). Additionally, the odor and hygiene influences of livestock manure have been considerable. Nitrogen and phosphorus losses Nitrogen leaching has caused groundwater contamination, eutrophication and also, in- directly, nitrous oxide emissions (Vitousek et al., 1997; Carpenter et al., 1998). Phosphorus losses contribute to the eutrophication of wa- terways. The level of nitrogen and phospho- rus leaching has been influenced by soil and weather conditions and by cultivation prac- tices (type, timing and amount of fertiliser application; crop type, timing of cultivation, and the type manure spreading technology (e.g. Shepherd et al., 2003). The overuse of fertilizer nutrients has increased the soil nu- trient reserves, as found in Finland for phos- phorus (Saarela, 2002). The amount of P lost to surface waters has been reported to increase with the P content of the soil (Shar- pley and Rekolainen, 1996). When meas- ures to reduce nutrient leaching are devel- oped, an important point is that more than 90% of the loading from arable fields to sur- face waters in Finland has been found to en- ter the water bodies outside the growing sea- son (Puustinen et al., 2007). Gaseous nitrogen emissions Agriculture is the main source of ammo- nia emissions. In Finland, 90% of ammonia emissions originate from agriculture (Grön- roos et al., 2009). Ammonia emissions from agriculture arise mainly from manure (ECE- TOC, 1994; Grönroos et al., 1998). Animal housing, manure stores and manure spread- ing are the major sources of ammonia emis- sions. In addition to odor emission, ammonia is the main acidifying pollutant from agricul- ture (OECD, 2013). Manure management is one source of nitrous oxide (N2O) emissions (Grönroos et al., 2009). N2O is a greenhouse gas which contributes to climate change and Table 6. Annual production of livestock manure (m3/animal/year) in 1990 and 2012 (Hellstedt et al., 2013). Present design instruction (RMO) for manure storages in Finland (RT MMM/MTH-20919). 1990 2012 RMO Manure type manure (m3/animal/year) Dairy cows slurry manure 18.3 25.8 24.0 Fattening pigs slurry manure 2.0 2.4 2.0 Sows and piglets slurry manure 9.1 9.3 7.0 22 MTT SCIENCE 29 can catalyze the destruction of ozone (Vi- tousek et al., 1997). It is produced in soils aerobically during nitrification and anaero- bically during denitrification. Emissions arise from the manure management and nitrogen applications (e.g. mineral fertilisers and ma- nure) to the soil (US-EPA, 2006). Hygiene Livestock manure can harbour a wide range of bacterial, viral, and parasitic pathogens (McAllister and Topp, 2012). Therefore, it can pose a microbiological risk. There has been increasing concern about the effects of pathogens possibly present in animal ma- nure on human and animal health (Bicudo and Goyal, 2003). In recent years, outbreaks of food‐borne diseases caused by enteric mi- croorganisms have received much attention, leading to increased consumer concerns about the safety of their food supply. A grow- ing concern has been that many previously unrecognized foodborne pathogens, includ- ing Campylobacter jejuni, Escherichia coli 0157:H7, and Listeria monocytogenes, have emerged during the past twenty years (New- ell et al., 2010). Adding to the problem, treat- ment is more difficult due to an increase in antibiotic resistance among common food- borne pathogens (e.g. Gilchrist et al., 2007). Moreover, it has been evaluated that new patterns of resistance to antimicrobial agents and new forms of virulence emerge in old pathogens and influence the risk of epidem- ics in the future (Velge et al., 2005; Martella et al., 2010). Some practices associated with intensified production have been estimated to have the potential to increase the risk of zoonoses (Liverani et al., 2013). The most important zoonotic pathogens and viruses in cattle and swine are listed in Tables 7 and 8. Costs due to animal diseases have normal- ly been associated with reductions in animal populations and production (IFAH, 2012). Even those diseases which are not classified as dangerous may result in significant costs. There are also costs related to the mitigation of disease, which include the money and re- sources expended to monitor, control and, in extreme cases, eliminate the disease agent (IFAH, 2012). Figure 2. Possible loss pathways of nutrient elements from the feed–animal–manure–soil–crop chain. Loss- es via gaseous emissions to the atmosphere are shown in the upper half, losses via leaching and runoff of nutrient elements to the subsoil and to groundwater and surface waters, and accumulation of nutrients into the soil, in the lower half (modified from Oenema et al., 2007). MTT SCIENCE 29 23 Table 7. Zoonotic pathogens of potential interest in cattle and swine. (Pell, 1997; Pelzer and Currin, 2007; Raizman et al., 2004; Hutchison et al., 2004, 2005; Bhaduri et al., 2005; Farzan, 2010; Esaki et al., 2004). Pathogen Common species Other remarks Campylobacter Campylobacter jejuni, Campylobacter coli Widespread in the intestinal tract of warm- blooded animals. One of the most often encountered bacteria responsible for human gastro-intestinal infections Cryptosporidium (Protozoa) Cryptosporidium suis, Cryptosporidium parvum Some Cryptosporidium spp. such as Cryptosporidium parvum have become a public health concern Escherichia coli E. coli O157:H7 A significant number of foodborne disease outbreaks have increasingly been attributed to Escherichia coli O157:H7 Giardia (Protozoa) Giardia duodenalis (syn. G. lamblia, G. intestinalis) Giardia duodenalis is a commonly identified intestinal parasite of mammals, including humans. Human giardiasis, most often associated with drinking water, is frequently diagnosed in the United States Leptospira The severe pathogenic serovars Pomona in pigs and Hardjo in cattle have not been reported in Sweden European Centre for Disease Control (ECDC) is pinpointing leptospirosis as an important infectious disease of particular interest in Europe. The zoonosis of leptospirosis, which is of worldwide distribution, is caused by different pathogenic serovars belonging to the genus Leptospira and is endemic in most tropical and temperate climates Listeria Listeria monocytogenes Mycobacterium Mycobacterium avium subsp. paratuberculosis (Mycobacterium paratuberculosis) M. paratuberculosis is the causal agent of paratuberculosis or Johne’s disease, which is a common and chronic intestinal disease. Salmonella Salmonella can be isolated from numerous animal species and is known to be a principal zoonotic bacterium causing symptoms such as diarrhoea, fever and septicaemia. Salmonella is a causative pathogen in food-borne illness. There is considerable information on antimicrobial resistance in Salmonella of human and food animal origin. Yersinia Yersinia enterocolitica Swine are the primary reservoir from which Y. enterocolitica strains pathogenic to humans are isolated Major bacterial food-borne pathogen by the pork production and processing industry in the United States The health risks associated with animal operations depend on various factors. The most important ones appear to be related to intensified systems with high animal densities and the concentration and vir- ulence of pathogenic microorganisms in animal manure (Bicudo and Goyal, 2003; Létourneau et al., 2010). The nature and concentration of zoonotic pathogens ex- creted by animals differ according to ani- mal species and health, nutrition, age and housing environment (Bicudo and Goy- al, 2003; Cliver, 2009; Létourneau et al., 2010). In manure, bacteria and parasites persist for a time depending on geographic location of the farm, on the physicochem- ical composition of the manure (tempera- ture, pH, free ammonia, and solids con- tent), on aeration and on handling and storage management (Jones, 1982; Bicu- do and Goyal, 2003; Topp et al., 2009; Létourneau et al., 2010). The ability of the pathogens to survive for long periods and through treatment to remain infective in the environment until ingested by human or animal hosts has been an added concern (Bicudo and 24 MTT SCIENCE 29 Goyal, 2003). Most bacteria can survive and even multiply in environments out- side the animal, such as in livestock ma- nure (Strauch, 1991). Viruses cannot mul- tiply outside the animal, but are capable of surviving for long periods of time (even months) in the environment, depending on environmental conditions (Spiehs and Goyal, 2007). Odors and other air emissoins Odors are an increasingly difficult and pressing problem for the agricultural in- dustry. They have been a major factor causing complaints and have had a par- ticular relevance for environmental permit issues concerning barn expansions or con- struction of new buildings (Arnold et al., 2006). The primary reason is the increas- ing unit size and concentration of large numbers of farm animals on a single farm, increasing the potential for nuisance and environmental problems including odor annoyances, greenhouse gases, ammonia and hydrogen sulfide emissions (Jacobson et al., 2001a). Dust (a combination of ma- nure solids, dander, hair, and feed), path- ogens, and flies from animal operations have also been airborne emission concerns (e.g. Jacobson et al., 2001a). A variety of potential health effects related to odorous air have been reported, including symp- toms such as irritation, headache, nausea, etc. (e.g. Schiffman et al., 2004). Air emis- sions also pose the threat of aerosol trans- missible diseases (Millner, 2009). All these problems have been forecasted to contin- ue to increase as suburban development encroaches upon areas that are primarily used for agricultural purposes (Schiffman et al., 2004). Livestock odors originate from three pri- mary sources: animal buildings, manure storage units, and land application of ma- nure (Jacobson et al., 2001a). Of these sources, land application of manure is probably the greatest source of odor emis- sions and complaints. The study of Ar- nold et al. (2006) demonstrated the im- portance of uncovered manure storage tanks for overall odor emissions from ani- mal houses. Approximately half of the de- tected pig farm odor emissions originated from the uncovered manure storage tanks and half were from the pig farm build- ings during the summer season. Tempera- ture had a strong influence on odor emis- sions, as during the winter season almost no odor emissions were detected from ma- Table 8. Some of the most important viruses affecting cattle and swine. (Martella et al., 2010; Bøtner and Balsham, 2012). Virus Affected animals Other remarks/aspects Foot-and-mouth disease virus Affects cattle, pigs and sheep One of the world’s most economically important diseases Classical swine fever virus (and the related bovine viral diarrhea virus) Affects cattle and pigs, is excreted in faeces by infected animals Normally absent from Europe, but serious outbreaks of the disease can occur Swine influenza virus Widely distributed in pig populations, causes a relatively mild, largely respiratory, disease Poses a threat to human health. Contains segments derived from different influenza viruses, with a potential for novel reassortment Porcine parvovirus Replicates within the gut and thus can be present in swine slurry Widely distributed, being endemic in most countries; causes reproductive failure in swine Rotavirus Rotavirus-associated enteritis is a major problem in young calves and in weaning and post-weaning piglets Poses a zoonotic potential, comparison of genetic sequences of human and animal rotaviruses have revealed close identity MTT SCIENCE 29 25 nure storage tanks (Arnold et al., 2006). Teye (2008) and Teye and Hautala (2008) introduced a theoretical ammonia emis- sion model and measured the ammonia emissions from dairy houses in Finland and Estonia. Ammonia emissions from dairy buildings varied between 0.04 and 0.58 g m-2 h-1. They concluded that the critical parameters which should be con- sidered in reducing ammonia emissions in dairy buildings were manure temper- ature, pH and total ammoniacal nitrogen of the manure. 1.2.3 Environmental legislation Nitrates Directive and EU agri-environmen- tal support The implementation of the Nitrates Direc- tive (91/676/EEC 1991) is one of the policy measures aimed at decreasing nitrogen dis- charges and losses from agricultural sourc- es. Different measures are in place in the various EU countries to meet this direc- tive. In Finland, the Nitrates Directive is transposed to national legislation through the Environmental Protection Act (4th Feb- ruary 2000/86, paragraph 11.6) and Gov- ernment Decree No 931/2000 (9 Novem- ber 2000), adopting the whole territory approach in the Nitrates directive instead of designating specific nitrate zones as in many other countries. The Decree contains provisions on good agricultural practices, storage of manure, spreading and allowable quantities of fer- tilizers and silage liquor, analysis and re- cording of nitrogen in fertilizers and en- forcement of the Decree (931/2000). In practice, current legislation restricts the spreading of slurry manure or urine in the autumn and prohibits spreading during the time period from November 15th to April 1st. Legislation also restricts the to- tal amount of manure permitted to be ap- plied per hectare according to its nitro- gen content. In parallel with the implementation of the Nitrates Directive, Finnish farmers have widely adopted the EU agri-environmen- tal support scheme (96% coverage of the cultivated area during the second period 2000–2006). Environmental support for basic and additional measures has been paid to farmers who have met the eligibil- ity criteria laid down in Government De- cree No 644/2000 (26 June 2001) and un- dertaken the basic measures related to the following activity areas for five years: en- vironmental planning and monitoring in farming, basic fertilization levels of arable crops, plant protection, filter strips, biodi- versity and landscape management. Environmental permit The Environmental Protection Decree (169/2000) and the Environmental Protec- tion Act (86/2000) regulate animal shelter permit requirements concerning the keep- ing of animals in production buildings. The environmental permit procedure es- timates impacts on ground and surface waters, adverse smell and emissions into the air, manure storage, transportation and field application. Manure spreading and arable farming are not licensed activ- ities. However, in license permissions ara- ble land area and its adequacy in relation to the number of animals is taken into account. 1.3 Manure treatment technologies Manure treatment has been under inten- sive study, and different treatment technol- ogies have been extensively reviewed in a number of recent publications and inven- tory descriptions (e.g. Foged, 2010, 2011; Burton, 2007; Vanotti et al., 2009; Schou- mans et al., 2010; Sommer et al., 2013). Large numbers of scientific publications have also dealt with manure treatment. In this section, manure treatment tech- nologies are first categorized coarsely into physical, chemical and biological methods to introduce briefly the processes involved in treatment technologies. Treatment tech- 26 MTT SCIENCE 29 nologies are further categorized into phos- phorus recovery-oriented methods, meth- ods to treat the liquid fraction (comprising mainly those technologies covering nitro- gen separation recovery), odor treatment technologies, and methods aiming at path- ogen reduction. 1.3.1 Method categorization into physical, chemical and biological processes Physical methods Physical methods include technologies in- volving the application of physical forces to treat the manure, e.g. solids-liquid sep- aration and/or the use of heat and pressure. Solids-liquid separation has been achieved through settling or by using mechani- cal methods (e.g. using screens, sieves or grates), including intensified separation us- ing drum filters, filter pressing, belt presses, screw pressing or centrifuges (e.g. Møller et al., 2000; Burton, 2007; Hjorth et al., 2010). In combination with chemicals to increase solid material flocculation, solids- liquid separation has been used more ef- fectively to remove nutrients from manure (e.g. Hjorth et al., 2008, 2010; Walker et al., 2010). Other physical methods include drying, incineration, pyrolysis, combus- tion, and gasification (e.g. Schoumans et al., 2010; UNEP, 2009). Chemical methods Principal chemical processes include chem- ical coagulation, precipitation, disinfection, oxidation, neutralization, stabilization, and ion exchange methods (e.g. McCabe et al., 2005). Chemicals have also been used to control odors (e.g. Ritter, 1981; Zhu et al., 1997) and pH (Petersen et al., 2012). Elec- trochemical methods are incorporated in chemical methods. These methods com- prise electrodialysis, electroflotation/co- agulation, and electrochemical oxidation methods (e.g. Laridi et al., 2005). Biological treatment Biological processes make use of natural- ly occurring microorganisms or/and added inoculants to degrade manure in the pres- ence of oxygen (aerobic) or in its absence (anaerobic). Manure treatment using bio- logical methods has been designed to re- duce the odor of the slurry, to reduce the slurry nutrient (e.g. nitrification-denitrifi- cation) and organic matter concentrations and for stabilization of the slurry, to reduce the content of pathogenic microbes in the slurry, and to produce energy (Ndegwa, 2003; Juteau et al., 2004; Park et al., 2005; Zhang and Zhu, 2006). The used biolog- ical methods include anaerobic treatment methods, e.g. anaerobic digestion (Nasir et al., 2012) and fermentation (Banister and Pretorius, 1998), and aerobic treat- ment methods, e.g. composting (Bernal et al., 2009). Sequential aerobic-anaerobic treatment methods, e.g. enhanced biolog- ical phosphorus removal (Toerien et al., 1990), have also been used. 1.3.2 Phosphorus removal and recovery technologies Phosphorus recovery technologies have been extensively reviewed e.g by Schou- mans et al. (2010). The methods relevant to the empirical work of the present study are briefly described in the following. Solid-liquid separation The simplest and most widely used meth- od to treat manure phosphorus is solid- liquid separation. Foged et al. (2011) de- scribed 10 mechanical, chemical and other technologies for active separation of slur- ries in their inventory, and estimated the amount of livestock manure treated by separation to be 3.1% of the entire live- stock manure production in the EU. The most used separation methods were separa- tion by drum filters, screw pressing, sieves, centrifuges, and natural settling (Foged et al., 2011). Most separation methods are based on particle size and particle density differ- ences (Burton, 2007; Hjorth et al., 2010). More sophisticated devices such as cen- trifuges and chemical-enhanced settling MTT SCIENCE 29 27 can achieve higher separating efficiencies, but involve additional equipment and/or management requirements (Hjorth et al., 2010). The efficiency of the separation also depends on the physical (e.g. DM content, particle size) and chemical composition of the animal manure, and manure type (Zhu et al., 2001; Møller et al., 2002; Pop- ovic and Jensen, 2012). Long storage time typically impairs the separation result by reducing the particle size of the solid ma- terial (e.g. Zhu et al., 2000). In general, low-solids swine manure settles more easily, whereas dairy manure with its higher solids content is more easily sepa- rated by mechanical separation (Møller et al., 2002). Mechanical separation meth- ods do not remove salts and have only had a limited capacity for separating out small (< 1 mm) particles with P, N and odorous compounds being present in high propor- tions in this particle class (Zhu et al., 2001; Popovic, 2012). Chemical methods Chemical precipitation or coagulation and flocculation with various salts of alu- minum, iron and other inorganic or or- ganic chemicals, and also lime, have been widely used to treat manure, aiming for pH change, coagulation of dry matter, and for P removal (e.g. Hjorth et al., 2008; Hjorth, 2010). Precipitation occurs if the concentration of a compound exceeds its solubility in a solution. Thus, precipitation is achieved by a) change in ion concentration of the manure, b) change in pH, or c) by using both of these methods (e.g. Wang et al., 2005). For example, following the addition of multivalent cations (Fe3+, Fe2+, Ca2+) to the slurry some phosphate will precipitate (Hjorth et al., 2008) due to formation of, for example, ferric phosphate (FePO4), fer- rous phosphate (Fe3(PO4)2) and calcium phosphate (Ca3(PO4)2). In addition to precipitation, the addition of multivalent cations to the slurry manure causes coagulation of the dry matter par- ticles of the slurry (Sherman et al., 2000; Hjorth et al., 2008). In coagulation, the multivalent cations neutralize or partial- ly neutralize a particle’s negative surface charge by adsorbing the oppositely charged ions to the particle surface, creating a dou- ble layer and thereby removing the electro- static barrier preventing aggregation; this process is termed ‘charge neutralization’ (Hjorth et al., 2010). The addition of polyelectrolyte polymers to slurry manure has been used to induce flocculation (Vanotti et al., 1996; Vanotti et al., 2002). Flocculation of small parti- cles into larger ones increases the effective particle size by creating flocs (aggregation of small particles and smaller flocs). These flocs separate out from the liquid phase of raw slurry more rapidly than their constit- uent small particles alone would do (Pe- ters et al., 2011) and they can be more easily retained by screens, or enhance sep- aration of colloidal particles by their set- tling (Vanotti et al., 2002). Cationic pol- ymers have been shown to be much more effective than anionic or neutral polyacryla- mide polymers in removing organic solids from swine manure wastewater (Vanotti et al., 1996), most probably due to slurry particles carrying negative charges (Grego- ry, 1989). However, aluminium has been shown to be much more effective than pol- yacrylamides in reducing slurry total solids (TS) and P (Sherman et al., 2000). When selecting chemicals to be used for precip- itation/coagulation, factors that need to be taken into account include: precipitate formed, solubility of the chemical, reac- tion time, and handling safety (e.g. Burns et al., 2003; Le Corre et a., 2005; Wang, 2005). Most salts of monovalent cations (e.g. Na+, K+) are water soluble, whereas divalent cations (Mg2+, Ca2+) salts are less soluble; solubility depends e.g. on crystal structure (Haynes, 2012). Trivalent ions (e.g. Al3+ and Fe3+), although efficient pre- 28 MTT SCIENCE 29 cipitants, form sparingly soluble Al and Fe phosphates, which are hardly available to plants (Dao et al., 2001; Hyde and Mor- ris, 2004). Struvite precipitation is a phosphorus recov- ery technology taking advantage of chemi- cal precipitation (Burns et al., 2002, 2003; Wang, 2005). Struvite is a crystalline pre- cipitate, magnesium ammonia phosphate hexahydrate (MgNH4PO4*6H2O). Stru- vite crystallization depends on factors such as component ion ratios, interfering ions, temperature, and solution pH (Wang et al., 2005). Struvite precipitation from manures has been typically achieved by adding Mg in a slurry (e.g. MgCl2, MgO), and/or by in- creasing slurry pH by addition of chemicals or by using aeration (Burns and Moody, 2002; Suzuki et al., 2002). Complex or- ganic solutions, such as manures, have re- quired higher than stoichiometric magne- sium additions to overcome complexing reactions (Burns et al., 2003). Struvite has been used as a valuable slow-release fertiliz- er (De-Bashan and Bashan, 2004). Biological phosphorus removal Enhanced biological P removal takes ad- vantage of the ability of polyphosphate- accumulating organisms (PAOs) to accu- mulate large quantities of polyphosphate within their cells (Shapiro, 1967). To in- duce biological phosphorus removal, an alternation between anaerobic and aerobic conditions is required, together with an ex- tracellular supply of energy in the form of easily degradable organics (Toerien et al., 1991). During the anaerobic phase, PAOs produce energy by hydrolyzing polyphos- phates, take up carbon source to produce polyhydroxyalkanoates, and release solu- ble orthophosphate or PO4 3--P (Petersen et al., 1998). Since poly-P is broken down to PO4 3--P for energy supply, the phosphate concentration in the anaerobic phase in- creases. The anaerobic phase needs to be followed by an aerobic phase. In this phase, polyhydroxyalkanoates are consumed to generate energy, and PAOs accumulate large amounts of phosphorus (the phos- phorus fraction of phosphorus-accumu- lating biomass is 5–7%) within their cells by reincorporating the PO4 3--P into new microbial cellular poly-P (De-Bashan and Bashan, 2004; Montag, et al., 2009). The polyphosphate-accumulating organ- isms have a strict requirement for cyclic anaerobic, anoxic and aerobic conditions which consequently makes the bio-P-re- moval process rather complex (Zuthi et al., 2013). Biological P removal has been wide- ly used for treating municipal and indus- trial wastewaters, but with manure slurries this technology is mainly in a developing phase. 1.3.3 Treatment of the liquid fraction after solid-liquid separation The liquid fraction obtained after sepa- ration contains soluble components (e.g. soluble ions and N compounds) and some colloidal and suspended particles (includ- ing part of the phosphorus and organic compounds), although the largest particles have been removed (Burton, 2007; Popo- vic, 2012). Composition of the liquid frac- tion after solid-liquid separation depends on the efficiency of the preceding separa- tion step. Almost all liquid fraction treat- ment methods require a pre-treatment step (e.g. prefiltration) in order to function ef- ficiently (Hjorth et al., 2010). Filtration Microfiltration (MF) and ultrafiltration (UF) systems are basically highly efficient solid-liquid separators, but have not been typically used for concentrating soluble elements (Hjorth et al., 2010). Microfil- tration retains particles between 0.1 and 5 µm, while UF removes macromolecules and particles in the 0.001−0.05 µm size range (Masse et al., 2007). Both processes deal with separation of suspended solids, colloids and bacteria, do not develop sig- nificant osmotic pressures and require low operating pressures (Hjorth et al., 2010). Microfiltration and ultrafiltration are usu- MTT SCIENCE 29 29 ally efficient in concentrating the nutrients associated with particles, such as phospho- rus, but for other constituents, e.g. ammo- nia and potassium, the retention requires nanofiltration (NF) and reverses osmo- sis (RO (Masse et al., 2007; Hjorth et al., 2010)). NF membranes retain most organ- ic molecules with a molecular weight great- er than 200–400 Da (Masse et al., 2007). NF can retain some soluble salts (charged molecules such as Ca2+, Mg2+, NH4+). Viau and Nomandin (1990) reported that a NF membrane (150 DA) operated at 2.1 MPa pressure retained 52% and 78% of TAN and potassium, respectively, in the effluent from an aerobic reactor treating pig ma- nure. The most recent technological de- velopment is to use gas permeable mem- branes to remove and recover NH3 from liquid manure taking advantage of mem- branes through which only NH3 can be transported (Vanotti and Szogi, 2010). Because of fouling, micro- and ultrafiltra- tion membranes can only be used to sepa- rate pre-treated slurry, e.g. runoff streams from centrifuges (Hjorth et al., 2010). The fouling problems have been even more se- vere for nanofiltration and reverse osmosis than for micro- and ultrafiltration. Nanofil- tration or reverse osmosis can therefore only be used for separation of dissolved compo- nents from the permeate produced by an ul- trafiltration unit (Hjorth et al., 2010). Reverse osmosis Reverse osmosis is a pressure-based mem- brane purification technique. The prin- ciple is based on the use of pressure to force a solvent (usually pure water) to pass through a semipermeable membrane from a solution of higher concentration towards one of lower concentration, i.e. the oppo- site direction to that dictated by osmo- sis (Kucera, 2011). The osmotic pressure is related to salt concentration and exter- nal pressure. Reverse osmosis membranes theoretically retain all dissolved salts and organics with a molecular weight greater than approximately 100 Da (Masse et al., 2007). Rejection of dissolved salts rang- es from 95% to greater than 99% (Masse et al., 2007). As the feed is being concen- trated, osmotic pressure builds up in the system and high operating pressures must be applied (Kucera, 2011). The amount of pressure required depends on the salt concentration of the feed water. The more concentrated the feed water, the more pres- sure is required to overcome the osmotic pressure (Kucera, 2011). Reverse osmosis requires extensive manure pretreatments to prevent fouling, maximize membrane life, and increase flux (Masse et al., 2007). Electrodialysis Electrodialysis (ED) is a membrane sepa- ration method in which salt ions from one solution are transported through ion-ex- change membranes to another solution un- der the influence of an applied electric po- tential difference in an electrodialysis cell (Ippersiel et al., 2012). Electrodialysis has been used in combi- nations with other technologies to recov- er and concentrate ammonia from swine manure (Mondor et al., 2008; Ippersiel et al., 2012). Mondor et al. (2008) used elec- trodialysis and reverse osmosis. The re- sults obtained by these authors suggested that, under the conditions of their experi- ment, a maximum total NH3 -N concen- tration of about 16 g/L could be reached with the ED system. The maximum TAN concentration achievable by ED was limit- ed by water transport from the manure to the concentrate compartment and by am- monia volatilization (17%) from the open concentrate compartment. Ippersiel et al. (2012) used electrodialysis coupled with air stripping. In this study electrodialy- sis was used to concentrate ammonia and then air stripping from the electrodialysis- obtained concentrate solution without pH modification was used to isolate the am- monia in an acidic solution. However, low concentrate solution pH (8.6–8.3) limited NH3 volatilization toward the acid trap. 30 MTT SCIENCE 29 Ammonia stripping Ammonia stripping comprises reducing the content of ammonia in a liquid phase by transferring it in a tower into the gas phase. The method is based on first in- creasing the pH of liquid manure for con- verting the ammonium nitrogen contained therein into ammonia (US EPA, 2000; McCabe, 2005). This is followed by driv- ing the liquid manure through a tower filled with a packing material, promot- ing the volatilization of NH3 . The liquid manure is fed into the tower at its top end while blowing air into the tower from the bottom, resulting in the desorption of am- monia (US EPA, 2000; McCabe, 2005). The ammonia gas separated into the gas phase is further passed into water or acid, in which the ammonia gas adsorbs into the liquid (Bonmati and Flotats, 2003). In the stripping column, a large surface area between the liquid and the gas phase is needed in order to maximize the area for ammonia desorption and the rapid trans- fer of ammonia to the gas phase. This is achieved by filling the stripping column with carriers which promote the flow of a liquid in a thin film, while still allowing a high air flow through the column (Srinath and Loehr, 1974). Mass transfer of ammonia into the air is also dependent on the difference in the concentration of ammonia in the liq- uid phase and in the air phase (US EPA, 2000). High air flow rate enables a con- stant low concentration of ammonia in the air, promoting the transition of ammo- nia into the air phase (Arogo et al., 1999). Difficulties related to ammonia air strip- ping from manures have been related to high manure solids content,that has caused blockage of the tower and requires effi- cient pretreamtment (Rulkens et al., 1998). Moreover, air stirpping requires high pH (e.g. Bonmati and Flotats, 2003), although due to the high buffering capacity of ma- nure the pH increase is difficult to achieve (Sommer and Husted, 1995). Biological methods The N contained in the liquid fraction of manure can also be removed biologically by fixing it into biomass. Of the biologi- cal N-removal systems only the algal and duckweed pond systems apply resource re- covery, by using the produced algae and duckweed (Mulder, 2003; Craggs et al., 2011). Nitrogen has also been used for the production of fungi, amino acids, bacteria and yeast (Rulkens et al., 1998). In algal ponds ammonia is assimilated into algal biomass according to the equation: NH3 +5CO2+2H2O→C5H7O2N+5O2 (9) The energy use in algal ponds is required for mixing and pumping. The highest val- ue of the N-load has corresponded with the lowest efficiency (Mulder, 2003). In duckweed ponds, 41–68% of the applied N load has been assimilated in duckweed (Alaerts et al., 1996). 1.3.4 Odor treatment technologies Odor control, targeting in buildings and facilities, in manure storage systems, and in land application, has been carried out using a variety of approaches. The odors emitted from manure storages have been controlled by covering the storage tanks with a sealed lid, floating straw, natural crust etc. (Bicudo et al., 2003). In animal buildings, odor emissions have been re- duced by filtering the air coming from the production buildings and facilities using bioscrubbers or bio-filters (Sheridan et al., 2002). Odor emissions have also been con- trolled by treating slurry itself using aera- tion (Ndegwa, 2003; Juteau et al., 2004; Park et al., 2005; Zhang and Zhu, 2006), by using anaerobic digestion (Powers et al., 1997) or other biological treatment meth- ods (Ndegwa, 2003), by using chemical or biological additives added to slurry ma- MTT SCIENCE 29 31 nure (Ritter, 1989; Zhu et al., 1997; Zhu, 2000) or through diet manipulation (Sut- ton et al., 1999). In order to eliminate or reduce odors asso- ciated with animal manures, several types of chemical or biological compounds have been used according to Ritter (1981): (1) masking agents that override the offen- sive odors, (2) counteractants chemical- ly designed to block the sensing of odors, (3) odor absorption chemicals that react with compounds in manure to reduce odor emission, (4) feed additives affecting ani- mal digestion, (5) oxidizing agents such as hydrogen oxide and sodium hypochlo- rite that chemically oxidize compounds, and (6) biological compounds such as en- zymatic or bacterial products that alter the decomposition so that odorous com- pounds are not generated. Several studies have demonstrated the limited success of these commercial odor treatment additives (Patni, 1992; Zhu et al., 1997; Zhu, 2000). According to Zhu (2000) this might be due to the complexi- ty of odorous compounds in manure. Rit- ter (1989) postulated that some additives were only able to eliminate one or two types of odorant and that if these were not the major odorants present, the prod- uct did not work. Adding oxidizing agent can increase dissolved oxygen levels with- in a short time, but the effect is only tem- porary. In the case of microbial additives, since most indigenous bacterial genera in swine manure are strict anaerobes, they can always outcompete any added aerobes if the manure is maintained in anaerobic conditions (Zhu, 2000). 1.3.5 Treatment methods for inactivation of manure pathogens Physical methods Physical treatments for the inactivation of manure pathogens have included irra- diation (UV, gamma and electron beam irradiation), heat treatment, drying, and photocatalytic inactivation (Martens and Böhm, 2009). Combination technologies, such as thermal drying, have also been used. The main factors influencing the inactivation of the relevant pathogens are the temperature, the exposure time and the water activity in the material (Böhm, 2008; Table 9). Low water activity increas- es the heat resistance of microorganisms (Martens and Böhm, 2009). Chemical methods Chemical methods used for inactivation of manure pathogens have included disin- fection by the addition of chemicals and/ or due to pH shift, and/or due to redox po- tential shift (Block, 2001). Chemical dis- infection of liquid manure has been per- formed with several substances (Table 10). Depending on their chemical nature, the chemical disinfection methods are catego- rized as alkalis, surfactants, halogens, coal and wood tar derivatives, and miscellane- ous disinfectants (Gaskin and Meyerholz). The hygienization result has been depend- Table 9. Physical pathogen inactivation methods: exposure temperature, time, and disinfection results of treatments. Modified from Martens and Böhm (2009). Treatment Temperature (˚C) Exposure time Result Pasteurization 70 60 min Vegetative bacteria, viruses of moderate heat resistance and all infectious parasitic stages will be inactivated Pasteurization 90 60 min In addition to the above, some heat sensitive spores will be inactivated Microwave heating 80 <1s Inactivation of vegetative bacteria, moderately heat resistant viruses, and parasites. Thermal drying No significant inactivation of pathogens 32 MTT SCIENCE 29 ent e.g. on the presence of organic and sol- id material in slurry, on the concentration of the used chemical, and on slurry pH (Jones, 1982; Turner and Burton, 1997; Jenkins et al., 1998). Biological methods Aerobic and anaerobic treatments carried out in the mesophilic or in the thermophil- ic temperature range have had an inacti- vating influence on pathogens (e.g. Mar- tens and Böhm, 2009; Table 11). Table 10. Chemical pathogen inactivation methods: disinfectant type, mechanisms of action, exam- ples of chemicals used, and further remarks. (Turner and Burton, 1997; McDonnell and Russell, 1999; Block, 2001). Disinfectant type Mechanisms of action Chemical Remarks Alkalis A pH of at least 11.5 causes the inactivation Lime (calcium oxide or quick lime) A pH higher than 9 will inhibit most bacteria, and some viruses, but long wet contact times are required Surfactants A chemical compound that lowers the surface tension of an aqueous solution, promoting wetting Soaps, quaternary ammonium compounds Mild disinfectants, primary value is in aiding the mechanical removal of contaminated organic material. Generally used to disinfect non-porous surfaces. Do not possess substantial viricidal, fungicidal or sporicidal properties and are generally used in final rinses after mechanical cleaning Alcohols Cause membrane damage and rapid denaturation of proteins and subsequent cell lysis Ethyl alcohol (ethanol), isopropyl alcohol (isopropanol), n-propanol Rapid antimicrobial activity against vegetative bacteria, viruses, and fungi, but are not sporidical Aldehydes Strong association with outer layers of bacterial cells, influence inhibitory actions Glutaraldehyde Broad spectrum of activity against bacteria and their spores, fungi, and viruses Halogens/ halogen releasing agents Highly active oxidizing agents and thereby destroy the cellular activity Iodine, bromide and chlorine, Sodium Hypochlorite Calcium hypochlorite Widely used for disinfectant purposes, high concentrations required where organic carbon levels are high Silver compounds Interaction with thiol (SH) groups, effects on enzymes, inhibition of growth, and cell division Silver nitrate, Silver sulfadiazine Ozone Viruses more resistant to ozonation than bacteria; high ozone levels are required when organic carbon levels are high Peroxygens Releases free oxygen rapidly/ producing hydroxyl free radicals Hydrogen peroxide (H2O2) Peracetic acid (CH3COOOH) Broad spectrum efficacy against viruses, bacteria, yeasts, and bacterial spores Phenols General protoplastic toxicity and membrane active properties Possess antifungal and antiviral properties Bis- Phenols & Halophenols Effects on the cytoplasmic membrane Triclosan, Hexachlorophenole and Chloroxylenol Bactericidal, but P. aeruginosa and many molds are highly resistant MTT SCIENCE 29 33 1.4 Needs for manure treatment and the importance of nutrient recycling Manure contains considerable amounts of nutrients useable in plant production. In Finland in 2011, the manure P content was estimated to be 17.5 million kg (8.8 kg P/ha arable land), of which 77% origi- nated from cattle and pigs (Ylivainio et al., 2014). Salo et al. (2007) estimated that the amount of nitrogen contained by livestock manure was 42−55 kg/ha in Finland dur- ing the period 1990−2005. At present, the manure is spread on fields situated close to the farms. However, re- gional concentration of production as well as increase in farm size has led to a situa- tion in which local arable land is not suf- ficient to receive all the manure produced in the region (e.g. Ylivainio et al., 2014). Manure application is limited either by its nitrogen or phosphorus content. Phospho- rus limitation is a result of elevated con- centrations of easily soluble phosphorus on farmland (Saarela, 2002) due to the earli- er overuse of P. The use of manure as a crop fertilizer has been limited by its unfavourable N:P-ra- tio, relatively low nutrient concentrations, and wide variations in the amounts of nu- trients and in other properties (Luostarin- en et al., 2011). Moreover, part of the ma- nure organic nitrogen in soil is mineralized at a later date, reducing the controllability of manure nitrogen in fertilization (Wha- len et al., 2001). The relatively low pro- portion of soluble nitrogen and the high amount of phosphorus in relation to the needs of plants mean that manure fertiliza- tion cannot be optimized in the same way as inorganic fertilizers. Often additional inorganic N fertilizers are needed to sup- plement manure nutrient supplied for veg- etation growth. The importance of P recycling has been well recognized due to limited P reserves (Gilbert, 2009). The manufacture of N fertilizers is an energy intensive and green- house gas emitting process, as nearly 1 m3 of natural gas is required per kg of an- hydrous ammonia (Noble Foundation, 2001). Thus, better manure nutrient recy- cling involves both economic and environ- mental aspects. Moreover, manure appli- cation has been found to be expensive, as large volumes of slurry with relatively low nutrient content and high water content must be transported for long distances to fields (Kässi et al., 2013). Soil compaction has been found to be a serious risk when Table 11. Biological treatment methods: exposure temperature and time, and their disinfection results. Modified from Martens and Böhm (2009). Treatment Temperature (˚C) Exposure time Result Aeration treatment of liquid manure, aerobic thermophilic stabilization (ATS) 50 55 60 23 h 10 h 4 h Vegetative bacteria, Viruses of moderate heat resistance and infectious parasitic stages will be inactivated Composting of solid manure 55 65 > 2 weeks 1 week Anaerobic treatment at mesophilic temperature range 30–40 Does not lead to reliable inactivation Anaerobic treatment at thermophilic range Above 53 Exposure time of at least 20 h Can be effective for inactivating vegetative bacteria, viruses with moderate resistance and infectious stages of parasites 34 MTT SCIENCE 29 applying manure using heavy machinery on wet soils (e.g. Arvidsson et al., 2003). In Finland, limited application time dur- ing the early spring further complicates this situation as the soils are often wet, and the need to empty the manure storage tanks during late fall cause difficulties if the fall is very wet. As a conclusion, there is a need to im- prove the recycling and utilization of ma- nure as a raw material and fertilizer. In particular, the need is for concentration of nutrients, for nutrient recovery technolo- gies, and for volume reduction. Further- more odor and lack of hygiene are severe problems worldwide (their importance is explained in more detail in section 1.2.2 Environmental impacts related to animal manure) and efficient and cheap treatment technologies are needed. Existing waste water treatment technologies may offer ap- proaches to reach the desired goal. The nutrient recovery and recycling in con- ventional waste water treatment technol- ogies has, however, been rather weak, in particular considering N and P recycling (e.g. Ashley and Mavinic, 2009). Anoth- er aspect is that manure is a very com- plex material in terms of its nutrient con- tents, physico-chemical characteristics and its microbiological composition (e.g. Som- mer et al., 2013). These specific slurry fea- tures should be taken into more detailed consideration when developing new treat- ment technologies. In fact, development of manure treatment technologies requires a comprehensive approach and fundamental level knowledge on the physico-chemical and biological properties of the manure. 1.5 Theoretical background of the Manure treatment system of the present study The purpose of this section is to provide a theoretical background for the manure treatment system developed and examined in the present study. First, reactor types are reviewed. The reactor types are relevant to how the physico-chemical properties can be contolled in a system and determine the stability of the system. Physico-chemical conditions determine the operation of the entire system. The basis of the decompo- sition and humifiication reactions relevant to the biological processes of the present- ed treatment system and their dependence on the aerobic-anaerobic state of the sys- tem are also discussed. 1.5.1 Reactor types The reactors used for waste water treat- ment have been commonly categorized based the operation pattern, hydrau- lic characteristics, unit operation occur- ring, and entrance/exit conditions (Reyn- olds and Richards, 1996; Seyrig and Shan, 2007). Common reactor configurations have included plug flow reactors (PFRs), completely mixed batch reactors (CM- BRs), and completely mixed flow reactors (CMFRs). In addition, a CMFR has also been referred to as a completely mixed re- actor (CMR) or continous stirred tank re- actor (CSTR) (e.g. Hayes and Mmbaga, 2013). CSTR and PFR have probably been the two most widely accepted reactor regimes used for waste water treatment (Reynolds and Richards, 1996). In water treatment, a typical reactor is a long, narrow chan- nel, long pipe or tubular, or a series of long channels, because it typically cinvolves ex- posure to the disinfectant of interest for a specific duration of time (Reynolds and Richards, 1996). The main characteris- tic of a plug flow reactor (PFR) is that no mixing occurs in the direction of flow; however, complete mixing is assumed within the cross-sectional area of the reac- tor. Water and all suspended flocs of bac- teria move with the same velocity along the tube reactor (Seyrig and Shan, 2007). CSTR runs at steady state with a contin- uous flow of reactants and products (Sey- rig and Shan, 2007). The feed assumes a uniform composition throughout the reac- MTT SCIENCE 29 35 tor; the exit stream has the same composi- tion as in the tank. In other words, some of the elements entering the CSTR leave it immediately, because product stream is be- ing continuously withdrawn from the reac- tor. On the other hand, others may remain in the reactor indefinitely, because all the material is never removed from the reac- tor at one time (Seyrig and Shan, 2007). In contrast to the CSTR, the PFR exhib- its a continuous decrease in substrate con- centration and an increase in bacterial con- centration in the direction of flow (Seyrig and Shan, 2007; Hayes and Mmbaga, 2013). Hydraulic performance of a CSTR reactor has been improved by increasing the number of CSTRs in series (Reyn- olds and Richards, 1996). CSTRs in se- ries have commonly been used in the bio- logical treatment of industrial wastewater (Abu-Reesh, 2010). 1.5.2 Degradation and recondensation processes involved in biological treatment Decompositon processes During decomposition processes, microor- ganisms degrade the raw material to syn- thesize new cellular material and to obtain the energy for these catabolic processes. Several chemical transformations take place as complex compounds are broken down into simpler ones and then synthe- sized into new complex compounds (Fig. 3; Tölgyessy, 1993; Rajeshwari and Bal- akrishnan, 2009). Before the microorgan- isms can synthesize new cellular materi- al, they require sufficient energy for these processes. The two possible modes of en- ergy-yielding metabolism for heterotroph- ic microorganisms are respiration and fer- mentation (Jurtshuk, 1996; Madigan, et al., 2003). Respiration can be either aer- obic or anaerobic (Jurtshuk, 1996; Madi- gan, et al., 2003; Black, 2012). Aerobic respiration, involving oxidation- reduction reactions and with molecular ox- ygen as the final electron acceptor, is more efficient, generates more energy, operates at higher temperatures, and does not pro- duce the same quantity of odorous com- pounds as anaerobic respiration (USDA/ NRCS, 2000; Jurtshuk, 1996). Aerobes have been found to be able to utilize a greater variety of organic compounds as sources of energy, resulting in more com- plete degradation and stabilization of the material (Jurtshuk, 1996). In anaerobic respiration, the microorgan- isms use electron acceptors other than O2, such as inorganic nitrates (NO3–), sulfates (SO42–), and carbonates (CO32–), to obtain energy (Jurtshuk, 1996). Their use of these alternative electron acceptors in the ener- gy-yielding metabolism has been found to produce odorous or undesirable com- pounds, such as hydrogen sulfide (H2S) and methane (CH4) (Jurtshuk, 1996; Black, 2012). Anaerobic respiration leads to the forma- tion of organic acid intermediates that tend to accumulate and are detrimental to aerobic microorganisms (Jurtshuk, 1996; Madigan, et al., 2003). Aerobic respiration also forms organic acid intermediates, but these intermediates are readily consumed by subsequent reactions so that they do not represent as significant a potential for odors as in anaerobic respiration (USDA/ NRCS, 2000; Madigan, et al., 2003). Fermentation is an anaerobic catabol- ic process, and does not require oxygen. During fermentation, energy is released from an organic compound and the final electron recipient is also an organic com- pound, such as lactic acid or alcohol (Ma- digan, et al., 2003). The fermentation met- abolic pathway only uses glycolysis and produces two molecules of ATP (Madi- gan, et al., 2003). Humification Humification is a partial form of micro- bial decomposition, which, however, may also involve condensation and polymeriza- tion of decay products (Tölgyessy, 1993). 36 MTT SCIENCE 29 To a considerable extent this is an oxida- tive process, resulting in the formation of high-molecular-mass nitrogen compounds with a remarkable abundance of aromat- ic components. The humification process in soil is generally very slow, taking years to complete (Tölgyessy, 1993). In com- post environments with controlled condi- tions, humification has been clearly faster (weeks) (e.g. Diaz et al., 2011). Accord- ing to Ndewga et al. (2007a), stabilization has been reached within days during slur- ry aeration. Different theories exist concerning humus formation (Stevenson, 1994). The essential part of the macromolecule is a ring struc- ture imparted by aromatic compounds. Various side chains and functional groups are connected to the ring. The rings are in- terconnected by side chains and function- al groups or simple bonds, so that they can polymerize to form larger molecules. The rings are formed in the humification pro- cesses from lignin components. These re- actions are catalyzed mainly by microbial enzymes, and as a result very stable poly- mers with complex structures are obtained that can be grouped into three different types: humic acids, fulvic acids and humin (Stevenson, 1994). Humic substances con- taining carboxyl and phenolate groups can form complexes with ions in a solution (Hayes and Swift, 1978). Rate of decomposition Decomposability of different materials varies. This has an influence on e.g. re- quired treatment time. Due to the com- plexity of slurry manure as a raw material, it contains materials with different levels of degradability. This complicates the defini- tion of the required treatment time. Sug- ars, water-soluble nitrogenous compounds, amino acids and simple proteins, lipids, starches and some of the hemicelluloses are decomposed first at a rapid rate, where- as insoluble compounds such as cellulose, some hemicelluloses and more complex (crude) proteins are decomposed less rap- Figure 3. Decomposition of organic compounds in aerobic and anaerobic cycles (modified from Rajesh- wari and Balakrishnan, 2009). This conversion is achieved through a series of reactions. These reactions serve not only to liberate significant quantities of energy, but also to form a large number of organic in- termediates that serve as starting points for other synthetic reactions. Several chemical transforma- tions take place as complex compounds are broken down into simpler ones and then synthesized into new complex compounds. 45 alternative electron acc ptors in the energy- ielding metabolism has been found to produce odorous or undesirable compounds, such as hydrogen sulfide (H2S) and methane (CH4) (Jurtshuk, 1996; Black, 2012). Anaerobic respiration leads to the formation of organic acid intermediates that tend to accumulate and are detrimental to aerobic microorganism (Jurtshuk, 1996; Madigan, et al., 2003). Aerobic respiration also forms organic acid intermediates, but the e intermediat s are readily consumed by subsequent reactions so that they do not represent as significant a potential for odors as in anaerobic respiration (USDA/NRCS, 2000; Madigan, et al., 2003). Fermentation is an anaerobic catabolic process, and does not require oxygen. During fermentation, energy is released from an organic c mpound and the final el ctron recipient is also an organic compound, such as lactic acid or alcohol (Madigan, et al., 2003). The fer entation metabolic pathway only uses glycolysis and produces two molecules of ATP (Madigan, et al., 2003). igure 3 Decomp sition of rganic compounds in aerobic and anaerobic cycles (modified from Rajeshwari and Balakrishnan, 2009). This conversion is achieved through a series of reactions. These reactions serve not only to liberate significant quantities of energy, but also to form a large number of orga ic intermediates that serve as starting points for other synthetic reactions. Several chemical transformations take place as complex compounds are broken down into simpler ones and then synthesized into new complex compounds. Humification Humification is a partial form of microbial decomposition, which, however, may also involve condensation and polymerization of decay products (Tölgyessy, 1993). To a considerable extent this is an oxidative process, resulting in the formation of high-molecular-mass nitrogen compounds with a remarkable abundance of aromatic components. The humification Aerobic process Carbohydrates, fats, proteins Nitrogenous, carbonaceous, sulphurous compounds Ammonia, CO2, H2S Nitrates, CO2, sulphur Nitrates, CO2, sulphates Anaerobic process Carbohydrates, fats, proteins Nitrogenous, carbonaceous, sulphurous compounds Organic acids, carbonic acid , H2S, CO2 NH3, acid, carbohydrates, CO2, sulphides NH3, CO2, humus, CH4, sulphides MTT SCIENCE 29 37 idly, and lignin, waxes and resins are de- composed only very slowly (e.g. McDon- ald et al., 2011). Acceleration of biodegradation by seeding The use of seeding is a common proce- dure e.g. with anaerobic digestion and ac- tivated sludge during the start up-phase. Anaerobic biodegradation of substrates re- quires microbial consortia with hydrolyt- ic and methanogenic activities (Gerardi, 2003a). Populations of anaerobic micro- organisms typically require a significant period of time to establish themselves to be fully effective, and so a common prac- tice has involved seeding with an adequate population of both the acid-forming and methanogenic bacteria (Gerardi, 2003a). Long start-up times have also been report- ed for wastewater treatment processes (Pi- juan et al., 2011). According to Pijuan et al. (2011), one of the main challenging is- sues for the aerobic granular sludge tech- nology is the long start-up time of the bi- ological process when dealing with real wastewaters. Typically, the formation of aerobic granules with nutrient removal ca- pabilities has taken several months (Pijuan et al., 2011). Regarding composting stud- ies, good results have been achieved by mixing the starting material with mature compost in order to supply microbes capa- ble of degrading the lignocellulosic mate- rials and to shorten the composting peri- od or increase the quality of compost. For example, Wang et al. (2011) reported that lignocellulose degradation preceded 57.7% faster in compost inoculated with a ligno- cellolytic fungus. 1.5.3 Physico-chemical factors involved in treatment processes In the present section the entire system controlling physico-chemical factors is dis- cussed. These factors are relevant when the system functioning is evaluated and adjusted. OPR and the sequence of redox reactions Oxidation-reduction potential (ORP) and dissolved oxygen (DO) have been wide- ly used in the control and monitoring of aeration processes (e.g. Ndegwa, 2007b; Zhu et al., 2005, 2008). An oxidation-re- duction process is basically achieved by electron(s) transfer from the substance be- ing oxidized (referred to as reductant) to the substance being reduced (referred to as oxidant). Numerous biological processes such as degradation of organic matter are essentially oxidation-reduction processes (Fig. 4, Goronszy et al., 1992). ORP ranges required for nitrification and denitrification are typically +50 to about +225 mV and indicate the presence of dis- solved oxygen (O2) (Gerardi, 2003b; WEF, 2007). An ORP reading of +225 to +400 mV indicates the presence of oxygen and nitrate (NO3–) (WEF, 2007). ORP read- ings in the range of +50 to −50 mV indi- cate that no free available dissolved oxy- gen is present and nitrate is present as an electron acceptor (Gerardi, 2003b). There should be no free DO present in this zone so a DO meter would read zero mg/L. ORP readings less than −50 mV indicate that there is no free oxygen or nitrate pre- sent, and the microorganisms would be utilizing sulfate (SO2-4) as an electron ac- ceptor for their energy requirements (The- odore and Gong, 2013). Ammonia equilibria Below are shown factors that affect ammo- nia volatilization from a system. Ammo- nia volatilization takes place at the liquid/ gas interface. The volatilization tenden- cy depends on the equilibrium constants (speed of reaction, assuming equilibrium). Ammonia nitrogen exists in aqueous so- lution as either ammonium ion or ammo- nia, according to the following equilibri- um reaction: NH4+ ↔ NH3 (aq) + H+ (10) 38 MTT SCIENCE 29 The equilibrium concentrations of NH4+ and NH3 are pH and temperature (T) de- pendent (Fig. 5), as given by Emerson et al. (1975) (II; equation 11) and Lide (1993) (II; equation 12): (11)= 1 + = 1 + 10 (11) + 9 × 10 = 1 + = 1 + 10 (11) + 9 × 10 (12) where [NH3 ] is the free-ammonia concen- tration, [NH3 + NH4+] is the total ammo- niacal nitrogen (TAN) concentration, [H+] is hydrogen ion concentration, and Ka is the acid ionization constant for NH3 . The acid dissociation constant (pKa) value for NH4+/NH3 in aquous solution (25 °C) is 9.25 (Lide, 1993). Carbonate equilibria Carbon dioxide concentration at the in- terface determines its volatilization ten- dency. Carbon dioxide (CO2) dissolves in water, producing carbonic acid (H2CO3). Equilibrium is established between the dissolved CO2 and H2CO3 (Poling et al., 2001): CO2 (l) + H2O (l) ↔ H2CO3 (l) (13) This reaction is kinetically slow. At equi- librium, only a small fraction (ca. 0.2−1%) of the dissolved CO2 is actually converted to H2CO3. Most of the CO2 remains as solvated molecular CO2. Carbonic acid is a weak acid that dissoci- ates in two steps (Lide, 1993): (14) H2CO3+H2O↔H3O++HCO3-pKa1(25 °C)=6.37 (15) HCO3-+H2O↔H3O++CO32-pKa2(25 °C)=10.25 Figure 4. Oxidation-reduction potential (ORP) and metabolic process readings for organic carbon oxi- dation (process number 1), polyphosphate development (2), nitrification (3), denitrification (4), polyphos- phate breakdown (5), sulfide formation (6), acid formation (7), and methane formation (8) (modified from Goronszy et al., 1992). MTT SCIENCE 29 39 Figure 5. Proportion of NH3 /NH4+ ions in a solution at different pH values and temperatures (10, 20, 20 and 40 ˚C). Calculated using equatations 11 and 12. 0 20 40 60 80 100 7 8 9 10 Pr op or tio n of N H 3 /N H 4+ pH 10NH3 20NH3 30NH3 40NH3 10NH4 20NH4 30NH4 40NH4 The equilibrium between H2CO3 and bicar- bonate (HCO3-) and carbonate (CO32-) ions is pH- and temperature-dependent. Using carbonate mass balance and acid equilib- rium expessions, the fractional amounts of all carbonate species can be found as a function of [H+] (Poling et al., 2001; Lide, 1993): = = = (16) (17) (18) where Ka1 is 4.3*10-7and Ka2 is 5.6*10-11 Figure 6 summarizes the carbonate sys- tem. The form of carbonate in the solu- tion is very dependent on pH (Chang, 2000). At a low pH, the carbonate is pre- sent as carbonic acid, and it can evapo- rate. However, at high pH carbonate is present as carbonate anions and can inter- act with the cations present in the solution to form soluble and insoluble carbonates (Brown et al., 2011). For example, if Ca2+ is present, limestone (CaCO3) is formed, if Mg2+ is present, MgCO3 is formed and if NH4+ is present, NH4HCO3 is formed. The formation of the crystals depends on the relative amounts of ions in solution and on solubility product constants. Since ammonia and inorganic carbon ions are present at high concentrations in slurry manure, NH4+ and HCO3- are the pre- dominant ions. Moreover, crystal forma- tion and solubility is also pH dependent. Henry’s law Volatilization is dependent on NH3 or CO2 concentration at the interface accord- ing to Henry’s law. Henry s` law states that at constant temperature the partial pres- sure of a given gas in air is directly propor- tional to the concentration of that gas in liquid, when the dissolved gas is in equi- librium with gas in the air (Chang, 2000). pgas=hC (19) where • pgas is the partial pressure of the gas (of- ten in units of atm) 40 MTT SCIENCE 29 • C is the concentration of a gas at a fixed temperature in a particular solvent (in units of M or mL gas/L) • h is Henry’s law constant (often in units of atm/M) The Henry’s law constant (h) depends on temperature (Chang, 2000). It increases with increasing temperature. Therefore, the higher the temperature, the higher the nu- merical value of the Henry’s Law constant and the lower the solubility of the com- pound in water, and the easier the com- pound is volatilized (stripped). The solubility constant value is equal to the inverse Henry’s constant value (Chang, 2000). Accordingly, the solubility of a gas in a liquid depends on the partial pressure of the gas in the air at the interface. The numerical value of Henry’s Law constant increases with increasing temperature, and thus the solubility of the gas in a liquid de- creases with increasing temperature (Fig. 7). H2CO3 CO32- HCO3- Figure 6. Proportion of carbonic acid/bicarbonate ion/carbonate ion (H2CO3 / HCO3- / CO32-) in a solu- tion at different pH values at 25°C. Modified from Chang (2000). Figure 7. Solubility of carbon dioxide (CO2) and ammonia (NH3 ) in water at one atmosphere (101.3 kPa) in different temparatures. Modified from the engineering toolbox (www.EngineeringToolBox.com). 50 Figure 7 Solubility of carbon dioxide (CO2) and ammonia (NH3) in water at one atmosphere (101.3 kPa) in different temparatures. Modified from the engineering toolbox (www.EngineeringToolBox.com). CO2 NH3 MTT SCIENCE 29 41 2 Objectives of the study In the present study, a new slurry ma-nure treatment scheme was proposed and examined. The treatment scheme comprised biological treatment of manure with decreased contents of solids and P in a specially designed reactor regime fol- lowed by ammonia separation by ammo- nia stripping. The general aim of the study was to in- vestigate the feasibility of slurry manure treatment conducted in the designed re- actor system. One research question was: how does moderate or only slight aeration affect slurry manure properties when con- ducted in a series of continuously fed aer- ated tank reactors? Limited aeration was applied in order to maintain N in am- monia form, thus preventing nitrate for- mation, avoiding abundant CO2 forma- tion and favoring humification processes. Both swine and dairy slurry manure were studied. Further, swine slurry manure was treated by ammonia stripping after biologi- cal treatment in order to separate N. It was studied whether biological treatment alone was sufficient to increase slurry manure ef- fluent pH to a level enabling N separation by ammonia stripping without the use of chemicals. The theoretical basis of each process step was also discussed. Focusing on the biological treatment pro- cess, the more specific objectives and hy- potheses of the study were to: 1. Examine the operational conditions, dy- namics and stability of the serial treat- ment system, and to study process effi- ciency of the system (III). 2. Investigate whether it is possible to effi- ciently increase slurry manure pH with- out turning ammonium to nitrate with aeration treatment or losing it in gaseous form, or losing a substantial fraction of the carbon during biological treatment (I, III, Supplemental data). 3. Investigate whether the slurry-precipitat- ing characteristics can be strengthened and how the system influences odor (I). The hypotheses were: the hydraulic reten- tion time of the serial system is short (days) and the system is stable; the limited aera- tion increases the pH of slurry manure but nitrate formation is still limited; and the biological treatment reduces the odor of slurry manure. Focusing on the hygienic state of the pro- cess and changes in microbial community composition during the manure slurry pro- cessing, the more specific objectives and hypotheses of the study were to: 1. Evaluate the used seeding preparation and seeding during the start-up phase. 2. Study the hygienic state of the process by measuring viable counts of bacteria (IV, Supplemantal data). 3. Study the changes in microbial commu- nity composition and to compare the community composition of treated slur- ry with that of untreated slurry manure and of the soil used as the origin of the inoculant (IV). 42 MTT SCIENCE 29 The hypotheses were: the use of seeding and feedback of treated effluent enhanc- es the functioning of the system; the bi- ological treatment improves the hygienic state of the slurry manure; and the micro- bial community of treated slurry manure is different from that of untreated manure. Focusing on ammonia separation by am- monia stripping, the more specific objec- tives and hypotheses of the study were to: 1. Investigate feasibility of the sequential stripping procedure: (i) increasing the swine slurry manure pH without use of chemicals and with part of the slur- ry buffer capacity removed, and (ii) in- creasing the slurry pH with chemical treatments in between the stripping cy- cles (II). 2. Investigate changes in swine slurry ma- nure buffer properties during stripping (II). 3. Evaluate ammonia (NH3 ) stripping of the biologically treated dairy slurry ma- nure and the effect of magnesium ox- ide (MgO) treatment on stripping per- formance (III). The hypotheses were: the amount of chemicals can be reduced by combining the biological and chemical methods to in- crease the pH; ammonia removal by stip- ping lowers the buffer capacity of slurry manure; and the ammonia stripping is one option to produce mineral-form N fertiliz- er from slurry manure. Focusing on overall applicability of the de- signed manure treating technology and on the achieved environmental benefits, the more specific objective of the study was to: 1. Evaluate different process stages from the physicochemical and agricultural end-use viewpoints (I, II, III, IV). The hypothesis was that the developed slurry manure treatment system can sep- arate P and N from the manure and that the nutrients are in plant avaibale form, and the system clearly reduces the harm- ful odor of slurry manure. MTT SCIENCE 29 43 3 Materials and methods The slurry manure characteristics and treatment systems used in this work are described in detail in articles I-IV. More- over, supplementary material is presented on swine manure treatment (Supplemental data). The supplementary data presents the results of one study which was conducted with swine slurry manure using biological aeration treatment equipment including eight tank reactors in seriers. The perfor- mance and hygiene aspects of the biologi- cal process are examined. 3.1 Properties of raw slurries Swine slurry manure was taken from the collecting pit of a growing-finishing farm (I, II). Raw dairy slurry manure was ob- tained from the MTT (MTT Agrifood Research Finland) dairy farm (III, IV). The characteristics of raw swine and dairy slurry manures are presented in Table 12. 3.2 Technical description of the treatment system and treatment processes The designed slurry manure treatment scheme consisted of a biological aeration treatment process conducted in a specially designed reactor regime followed by ammo- nia separation employing ammonia strip- ping (Fig. 8, I and II). Before the aeration treatment, manure dry matter and P con- tent were decreased by separation (A in Fig 8; I and III). The initial stage of the treat- ment was a biological process, in which the system was filled with a microbial inocu- lum produced from soil-water suspension (B in Fig. 8, I:3.1). During biological treat- ment, slurry manure was continuously fed to the serial treatment system and main- tained under limited aeration to keep N in ammonia form, thus preventing nitrate formation, avoiding abundant CO2 forma- tion, and instead favoring humification pro- cesses. Feedback effluent from the last tank was used to inoculate the first tank (Fig. 9). Feedback also functioned as a buffer mech- anism against process interferences (III). Increased pH facilitated the subsequent N separation with ammonia stripping when the treatment was continued with N sepa- ration (C in Fig. 8; III). 3.2.1 Pre-treatment processes Prior to aeration treatment, part of the slur- ry manure solid content was removed by sedimentation or by mechanical means us- ing separation (step 1 in Fig. 9). Swine slur- ry manure was separated by natural un- forced sedimentation in an intermediate tank before transferring the liquid phase to a serial treatment system (I, II, Supplemen- tal data). Dairy slurry manure was mechan- ically separated (Bauer separator s650) into liquid and solid phases (III, IV). Table 12. Mean dry matter (DM) and chemical composition of raw slurry manures before sedimentation or mechanical separation (I:Table 1, III:Table 2). TOC, total organic carbon; Ntot, total nitrogen; NH4+-N, ammonium nitrogen; Ptot, total phosphorus. Manure type DM TOC Ntot NH4+-N Ptot K Mg Ca Na (%) (mg/kg (fresh weight)) Swine slurry(I) 1.85 3930 2255 1649 451 1188 203 635 668 Dairy slurry(III) 4.40 2190 1240 400 44 MTT SCIENCE 29 Figure 8. Schematic representation of the designed manure treatment sheme including solid separation step (A), biological aeration treatment equipment (B) and ammonia stripping equipment (C). For a more detailed technical description of A and B see Figure 9 and for C see publication II. AIR C Manure separation AIR A B 3.2.2 Serial treatment system Biological aeration treatment Both swine and dairy slurry manure aer- ation was conducted in pilot scale in a se- ries of continuously fed aerated tank re- actors consisting of six 600 l treatment tanks (hydraulic volume 2.96 m3 in I and II; 2.40 m3 in III and IV) connected in series by polyvinyl tubing (65 mm in in- ner diameter) (I; step 2 in Fig. 9). Before start-up the treatment reactors were filled with seeding material (an inoculant) and the slurry manure treatment was initiated with small amounts of slurry fed into the system (20−50 l/day). The feed-in flow in each tank was to the bottom of the tank and the outflow was through the crosswise upper corner in relation to the input flow. The surface effluent level remained con- stant when slurry manure ran gravitation- ally from one tank to another (I; Fig. 9). Recirculation (feedback) of effluent slur- ry from the last tank to the first tank was also introduced (step 3 in Fig. 9). One study was conducted with swine slur- ry manure using biological aeration treat- Figure 9. Components of the biological aeration treatment equipment. Prior to feeding, solids sep- aration is performed (1.). Aeration equipment con- sists of six aerated tank reactors (á 600 l) con- nected in series (2.). Part of the effluent slurry is feedback from the last tank to the first tank of the serial system (3.) (I). 5 3 2 1 Manure solid separation 2 1. 3 5 6 7 8 2. . A IR 3. . A IR AIR MTT SCIENCE 29 45 ment equipment including eight tank reac- tors in series. The results of this study are present as supplementary data. In the first study (I), the aeration equip- ment was the most primitive. Perforated tubes were used for aeration and mixing, air pressure was produced by a compressor and a pressure meter controlled the feed- in air pressure to each tank, which was ad- justed by a pressure regulator. In the other studies (II, III, IV and supplemental data), aeration was performed using membrane diffusers and air flow was produced by a high pressure blower and adjusted by man- ual rotameters. The treatment tanks were equipped with mechanical foam breakers, and gases released during processing were collected separately from each tank via an outlet duct. Dairy manure treatment differed from swine slurry manure treatment systems in that feedback was NH3 -stripped solu- tion after biological treatment (Fig. 8, III). Thus, in the dairy manure treatment sys- tem, slurry was transferred from the sixth process tank by gravitation to an addition- al overflow container, from which it was transferred with a submersible pump to a 1 m3 plastic container (Fig. 8). From this container the slurry was then transferred with a hose pump into a stripping tower. After stripping, effluent was transferred to a container from which a part was con- tinuously fed into the first treatment tank (feedback) (Fig 8; III: Fig. 1). Ammonia stripping The stripping tower was constructed using polyethylene plastic tubing (3.2 m height × 0.4 m internal diameter) (Fig. 8; II: Fig.1). The tower was randomly packed with 1.5 m3 of plastic rings, giving a specific area of 126 m2 of the filling material. Stripping was conducted with continuous forced air flow from the bottom of the tower. Slur- ry manure was continuously pumped into the top of the tower and distributed there evenly (II). A circulating warm water sys- tem was used to warm up the incoming liquid and air. The tower was insulated to maintain the desired temperature. The continuous effluent exiting the tower was collected in a closed container. Air con- taining NH3 was collected in water in a wet washer (II). Air stripping experiments were performed with a liquid flow rate of 20.83 L h-1 and an air flow rate of 20.83 x 103 L h-1. Thus, the liquid/air ratio was ad- justed to 1:1000 (II, III). 3.3 Experimental designs The functioning and feasibility of the proposed slurry manure treatment sys- tem was examined by carrying out con- trolled experiments (I, II, III, supplemen- tary data) and by analyzing the microbial community composition (IV, supplemen- tary data). Experimental designs of each study are presented in Table 13 and the reseach focus in each study in Table 14. Biological treatment experiments consist- ed of study periods of variable lengths during which the treatment system was loaded with variable feeding rates (I, III). Air stripping experiments included pre- tests using a pure ammonium reference solution with different concentrations, with or without warming the feed-in liq- uid and air, in order to test the general function of the stripping tower and opti- mal treatment parameters (II). Thereaf- ter stripping experiments were continued with slurry manure (II, III). 3.4 Analytics Methods used in the thesis are summa- rized in Table 15. Detailed descriptions of methods can be found in the original pub- lications I-IV. 3.4.1 Chemical methods and buffer capacity The measurements of process conditions and control (Table 15) were carried out daily at a depth of approximately 10 cm. Of the nutrients the concentrations of K, 46 MTT SCIENCE 29 Ta b le 1 3. E xp er im en ta l d es ig ns a nd p ar am et er s fo r th e b io lo gi ca l t re at m en t i n ea ch s tu d y (I− IV , s up p le m en ta ry m at er ia l). M an ur e Pr e- tr ea tm en t St ud y pe ri od H R T (d ay s) Lo ad in g (L /d ) Fe ed ba ck ra ti o A er at io n in te ns it y St ud ie d pa ra m et er s A rt ic le Sw in e se di m en ta tio n 1* 7 da ys 1* 21 d ay s 4. 3 73 4 & 9 78 0. 8± 0. 2: 1 0. 7– 7. 2 m g/ L D O D O , p H , T S, V S, C O D ,V FA , o do r, co lo r I Sw in e se di m en ta tio n 72 d ay s 6. 3 50 6 1. 0 1. 1– 7. 2 m g/ L D O -2 16 –( + )2 79 m V D O , O R P, pH , T S, TO C , h yg ie ne Su pp l. m at er ia l Sw in e se di m en ta tio n 9 33 0 1. 0 O R P, pH , T S, TO C ,V FA s II D ai ry m ec ha ni ca l se pa ra tio n 4* 30 d ay s 6. 5– 96 50 to 3 50 1. 5– 1. 6 -1 72 –( +) 30 8 m V O R P, pH , T S, T O C II I D ai ry m ec ha ni ca l se pa ra tio n 4* 30 d ay s 43 , 4 4, 4 8 & 6 .7 50 to 3 50 1. 5– 1. 6 -1 72 –( +) 30 8 m V O R P, pH , T S, V S hy gi en e, 16 Sr R N A IV H RT , h yd ra ul ic re te nt io n tim e; D O , d iss ol ve d ox yg en ; T S, to ta l s ol id s; V S, v ol at ile so lid s; C O D , c he m ic al o xy ge n de m an d; V FA , v ol at ile fa tty a ci ds ; O R P, ox id at io n re du ct io n po te nt ia l; TO C , to ta l o rg an ic c ar bo n MTT SCIENCE 29 47 Ta b le 1 4. R es ea rc h fo cu s in e ac h st ud y (I− IV ). M an ur e St ud y fo cu s St ud ie d pa ra m et er s A rt ic le Sw in e Bi ol og ic al tr ea tm en t p er fo rm an ce Fa te o f p H , c ar bo n (T C , I C , T O C ) a nd n ut rie nt s ( N to t, N H 4+ , N O 3– , K + , M g2 + , C a2 + , N a+ ) d ur in g bi ol og ic al tr ea tm en t I Sw in e H um ifi ca tio n In flu en ce o f b io lo gi ca l a er at io n tre at m en t o n slu rr y co lo r a nd p re ci pi ta tin g ch ar ac te ris tic s I O do r Fa te o f V FA s a nd o do r i nt en sit y du rin g bi ol og ic al tr ea tm en t I Sw in e H yg ie ne C ha ng es in e nt er ic in di ca to r o rg an ism s i n slu rr y du rin g bi ol og ic al tr ea tm en t Su pp le m . m at er ia l Sw in e Am m on ia st rip pi ng a nd b io lo gi ca l p H in cr ea se Ev al ua te th e ab ili ty to re m ov e am m on ia fr om b io lo gi ca lly tr ea te d slu rr y m an ur e w ith ou t c he m ic al p H in cr ea se II Se qu en ce st rip pi ng Ev al ua te th e re pe at ed c yc le s o f s tr ip pi ng o n N H 3 re m ov al II Sw in e Bu ffe r c ap ac ity Ev al ua te c ha ng es in b uf fe r p ro pe rt ie s d ur in g str ip pi ng II D ai ry Pr oc es s d yn am ic s a nd tr ea tm en t s ta bi lit y, lo ad in g in cr ea se fr om 50 to 3 50 L d -1 , f ix ed a er at io n, v ar yi ng T S lo ad Ef fe ct o f l oa di ng o n ox id at io n– re du ct io n po te nt ia l ( O R P) , t ot al so lid s c on te nt (T S) an d pH II I D ai ry Tr ea tm en t e ffi ci en cy Fa te o f c ar bo n (T C , I C , T O C ) a nd n ut rie nt s ( N to t, N H 4+ , N O 3– , K + , M g2 + , C a2 + , N a+ ) d ur in g bi ol og ic al tr ea tm en t II I D ai ry St rip pi ng p er fo rm an ce St rip pi ng p er fo rm an ce a nd th e ef fe ct o f M gO tr ea tm en t o n str ip pi ng e ffe ct iv en es s II I D ai ry H yg ie ne C ha ng es in e nt er ic in di ca to r o rg an ism s i n slu rr y du rin g bi ol og ic al tr ea tm en t IV D ai ry M ic ro bi al c om m un ity c om po sit io n C ha ng es st ud ie d by 1 6S rR N A ge ne se qu en ce a na ly sis in d iff er en t t re at m en t t an ks of th e sy ste m a nd c om pa ris on to th e co m m un ity c om po sit io n of u nt re at ed sl ur ry an d so il (th e or ig in o f t he u se d in oc ul an t) IV T C , t ot al c ar bo n; IC , i no rg an ic c ar bo n; T O C , t ot al o rg an ic c ar bo n; N to t, to ta l n itr og en ; V FA , v ol at ile fa tty a ci ds ; M gO , m ag ne siu m o xi de 48 MTT SCIENCE 29 Mg, Ca, and Na ions were analyzed with an atomic absorption spectrophotometer. Ptot and Ntot, after oxidation with peroxo- disulfate and NH4+ -N, NO3– -N, were an- alysed colorimetrically using a QuikChem Auto Analyzer (I−IV, supplementary data). Buffering curves (Table 15) were obtained by titrating the treated swine slurries against 1.0 mol L−1 HCl and 1.0 mol L−1 NaOH. To evaluate pH responses to in- creasing rates of base added, curves of the amount of base added vs. pH were plot- ted (II). Another buffer experiment was performed with corresponding congruent samples after carbonate exclusion. Carbon- ates were removed by acidifying the sam- ples to pH 4 (II). 3.4.2 Microbiological methods Selection and enrichment of initial microbi- al populations At the installation stage, the tanks were filled with effluent (sub-cultured inocu- lant). The procedure of sub-culturing is described in article I. Enumeration of enteric indicator organisms in slurry samples The hygienic state of the process was mon- itored by measuring viable counts of bac- teria, including total counts and counts of enteric indicator microorganisms includ- ing fecal coliforms, E. coli/coliforms, fecal streptococci and sulfite-reducing clostrid- ia (IV, supplementary data). Nucleic acid methods Nucleic acid methods included DNA ex- traction, PCR amplification of bacteri- al 16S rRNA, plasmid isolation and se- quencing and analyzing of the sequence data (IV). 3.4.3 Data analysis and statistics Statistical evaluation of the influence of different treatment parameters on slurry chemical composition (e.g. solids content, pH, nutrient contents, enteric counts) dur- ing biological treatment and the sequen- tial stripping procedure were determined using the MIXED procedure of SAS (Lit- tell et al., 1996) (I,II,III, IV, supplementa- ry data). Pairwise means comparisons were performed using the mixed default state- ment (t-test). Linear regression analysis of SAS was used to evaluate the influence of tank (of a serial system consisting of six treatment tanks in series) on odor, TS, VS, TOC, COD, VFA nutrient contents and enteric counts (I, IV, Supplemental data). MTT SCIENCE 29 49 Table 15. The methods used in this study. Symbols as in Tables 13 and 14. Method Described and used in Process monitoring, process control pH I, II, III, IV Dissolved oxygen (DO) I Redox potential (ORP) II, III, IV Chemical methods Solids (TS, VS) I, II, III, IV (VS I) Carbon (TC, IC, TOC) I, II, III Chemical oxygen demand (COD) I Volatile fatty acids (VFA total) I, II Individual VFAs II Odor intensity I Color change I Buffer capacity Titration curve II Nutrients Ntot, NH4+, Ptot, PO4 3-, I, II, III, IV Ca2+, Mg2+, K+ ,Na+ Microbiological methods Hygienic indicator microbes IV Total bacterial counts IV DNA extraction IV 50 MTT SCIENCE 29 4 Results 4.1 Effects of pre-treatment on manure composition Swine and dairy slurry manure had inher- ently different dry matter contents, and further differences in physicochemical composition led to differences in process feed contents between these manure types after pre-treatment by partial solids separa- tion (Table 16; I and III). Swine slurry ma- nure solids were removed using natural set- tling. TS removal efficiency reached 70% after settling (I). At the best, more than 80% of the swine manure total P and near- ly 90% of the phosphate phosphorus was removed before biological treatment (I). Dairy slurry manure was separated with a screw press. TS removal efficiency was 33% after separation (III). However, less than 5% of Ntot and only 12.5% of total phosphorus was removed with mechani- cal separation (Tables 12 and 16). In the case of dairy slurry slight sedimentation was observed after mechanical separation (III). 4.2 Biological treatment 4.2.1 Operational conditions, dynamics and stability of the serial treatment system All the experimental runs were conducted at room temperature. The temperature in each treatment reactor was 15–24˚C and no significant temperature increase due to aeration was observed between the sequen- tial tanks (I, III). To measure the oxidative state of the aer- ated tanks, DO and/or Redox-values were measured in each treatment tank. The DO content of swine slurry manure varied from 0.6 to 7.2 mg/L (six tanks; I: Table 2) and from 1.1 to 7.2 mg/L (eight tanks; Fig. 10, supplementary data). ORP levels ranged between -200 and +300 mV in the treatment tanks (Fig. 10, supplementary data; III: Fig. 2). Pattern of DO and ORP in a serial system The changes in ORP and DO with time in different treatment reactors (swine slurry manure Fig. 10, supplementary data; dairy slurry manure III: Fig. 2) were used to con- trol and monitor the aeration processes. It was found, as in other studies (e.g. Nde- gwa et al., 2007b), that ORP was a better indicator than DO to monitor the pro- cess. With respect to ORP values, a cer- tain pattern was easily detected: in general the lowest value was detected in raw slurry, which then increased from one tank to the next, reaching its highest level in the last tank (III: Fig. 2). The observed DO val- ues were erratic (Fig. 10; I: Table 2). Raw slurry was clearly distinguishable from the aerated tanks, but differences between the tanks or the relative order of the tanks were difficult to determine solely on the basis of the obtained DO-values. Defining optimal operation parameters based on ORP-values The ORP-values provided a useful tool to monitor and evaluate the process state. ORP changes were observed with a delay in subsequent tanks of the treatment sys- tem in the case of both increase and de- crease in ORP (III: Fig. 2). A systematic drop in redox value from one tank to the next was an indicator of process failure. The feed TS content appeared to contrib- ute to a rapid decrease in ORP values in the first and thereafter in the subsequent tanks as a result of increasing supply (III: Figs. 2 and 3). MTT SCIENCE 29 51 Ta b le 1 6. C he m ic al c om p os iti on o f th e sl ur ry m an ur e (± s ta nd ar d d ev ia tio n (S D )) af te r so lid s se p ar at io n p re -t re at m en t. R aw m an ur e p ro p er tie s ar e p re se nt ed in T a- b le 1 2 an d sy m b ol s in T ab le s 13 a nd 1 4. D M pH N H 4+ -N N to t P O 43 - P t ot T C IC T O C K + C a2 + M g2 + N a+ Pa pe r (% ) (m g/ L) Sw in e slu rr y I 0. 55 ±0 .2 5 13 67 ±4 15 09 ±2 2 21 ±6 82 ±4 26 26 ±8 2 15 43 ±9 4 10 83 ±1 65 96 4± 20 16 0± 11 49 ±4 36 1± 6 II 1. 01 ±0 .6 2 7. 9± 0. 25 14 94 ±1 80 18 16 ±3 72 50 ±1 6 25 2± 40 3 27 74 ±1 71 13 02 ±1 02 14 71 ±7 8 95 2± 19 6 20 4± 79 82 ±1 82 28 4± 53 D ai ry sl ur ry II I 2. 94 ±0 .0 1 12 20 ±1 0 21 90 ±2 0 35 0± 10 II I ( II I* ) 2. 42 7. 6 12 96 ±1 42 21 69 ±2 00 15 5 31 0± 53 11 02 3 17 01 93 22 ±6 40 28 01 81 9 35 3 51 6 II I ( IV *) 2. 16 ±0 .5 1 7. 9 85 9± 71 15 90 ±3 44 14 2 23 0± 11 1 64 96 11 51 53 35 ±7 16 44 2 21 8 21 08 19 7 *S tu dy p er io ds II I a nd IV in a rt ic le II I. 52 MTT SCIENCE 29 The obtained results suggest that a feed containing less than 1% TS content of swine slurry manure allowed a serial sys- tem loading with a maximum of 978 L/ day, corresponding to 3 days HRT-value of the treatment system with the used aer- ation rates (I). Higher solids content of the feed decreased the maximum loading vol- ume. Dairy slurry manure with less than 2% TS content allowed the system to be loaded with up to 350 L/d, corresponding to 6.8 days HRT-value (III). The obtained results suggest maintaining the redox-values of the treatment tanks ideally as follows: 1st tank at zero or slight- ly above, 3rd tank at around 100–180, and the 6th tank at 180–200 mV. Corrective measures should be initiated no later than when the ORP has fallen below zero in the third tank of the serial system (I and III). 0 2 4 6 8 1 5 15 25 35 45 55 65 69 D O (m g/ L) Day 0 1 3 6 8 -400 -200 0 200 400 1 5 15 25 35 45 55 65 69 R ed ox (m V) Day 0 1 3 6 8 Figure 10. Comparison of dissolved oxygen (DO) and Redox values during biological aeration treat- ment of swine slurry manure in a series of continuously fed aerated tank reactors. Measurements were made in untreated manure (0) and in individual tanks (1, 3, 6 and 8) of the serial treatment system (sup- plementary data). Effect of aeration on pH The average pH of pre-treated slurry ma- nure before aeration was between 7 and 8 irrespective of slurry type (Table 17). How- ever, after biological treatment the pH in- creased above 8.5 and even an average val- ue of 8.8 was reached in studies II, III. In some of the studies individual measured pH-values reached almost 9.0 (e.g. II: Ta- ble 2). The pH-increase due to aeration treatment appeared to depend on process loading. As shown in Table 17, lower av- erage pH-values were reached with short- er hydraulic retention time values (I: Ta- ble 3) or with higher carbon loading (III: Table 3). On the other hand, long hydrau- lic retention time increased the ammonia volatilization from the treatment tanks, which resulted in reduction of pH-values in the last tanks of the serial tanks system (III: Fig.4). MTT SCIENCE 29 53 4.2.2 Physico-chemical changes due to treatment Effect of treatment on solids and carbon content There was a clear decrease in the TS levels of slurry manure due to the treatment (Ta- ble 17). Effluent from the last tank con- tained on average 10, 31 and 16% less TS compared to slurry before aeration in the three experiments with swine manure (Ta- ble 17). Therefore, the higher the initial TS concentration was, the more TS content of swine manure was decreased due to aera- tion treatment. In the case of dairy slurry manure the ef- fect was less marked. The average TS re- duction percentage was 21% during peri- od III and 20% during period IV (Table 17; III). Carbon reduction during aeration treat- ment depended on initial carbon con- tent and varied between studies from 11 to 57% (Table 17). Swine slurry manure treatment led to an average of 13% TC and 11% TOC content reduction (efflu- ent from the last tank compared to feed- ing) in the first study (I: Fig. 2C) when the TS content of the feed was very low. When the feed TS concentration was higher than in paper I, 29% TC and 29% TOC re- duction in the second study (II: Table 1) and similarly 25% TC and 26% TOC re- duction in supplementary data (Table 17) were recorded. During dairy slurry manure treatment, ef- fluent from the last tank contained on av- erage 44–49% less TC and 47–57% less TOC compared to slurry before aeration in the two periods studied (Table 17; III: Tables 2 and 3). Greater carbon reduction was obtained with slurry having a lower carbon content (III: Period IV), when the greatest change in the amount of carbon occurred in the first treatment tank. Lower carbon content led to a 50% higher TOC reduction between tanks 1 and 6 (III). Odor The odor generation potential of the swine slurry was evaluated based on the VFA lev- els in the slurry. In addition, odor inten- sity was evaluated by a human panel (I). The biological treatment of slurry manure significantly reduced the odor intensity compared to that of untreated manure (I: Fig. 4B) when measured by sensing. The untreated slurry manure was evaluated as having a very strong odor, whereas in the last treatment tank no odor or only a very faint odor was reported (I: Fig.4B and Ta- ble 4). The best result was obtained with a longer hydraulic retention time (HRT 4 days). The odor intensity reduction due to biological treatment was only evaluated in the first study (I) with swine slurry, but similar observations (very effective odor re- duction) were also made during the other studies (II, III and IV). The VFA concentration was a less sensi- tive measure of the odor than the evaluat- ed odor intensity. The VFA concentration in swine slurry before aeration varied be- tween 404 and 484 mg/L (I: Fig. 4A) and 799±235 mg/L (II: Table 1). After treat- ment the VFA concentrations varied be- tween 330 and 484 mg/L (I: Fig. 4A) and 525±27 mg/L (II: Table 1) and were much higher than the level (below 230 mg/L) regarded as representing manure without an offensive odor, although odor intensity was reduced to a level of no odor or only a very faint odor. Color Swine slurry color change and darkness in- tensification from one tank to the next was observed visually and also confirmed by absorbance value comparison as an indi- cator of humification (I). The absorbance values were increased from 2.244 for un- treated slurry to values ranging from 3.079 to 4.000 in the last tanks of the treatment system (I). The peaks were observed at wavelength 382–383 nm (I: Fig. 6). 54 MTT SCIENCE 29 Ta b le 1 7. A ve ra ge f ee d co nc en tr at io n (F ee d ), ef flu en t co nc en tr at io n (E ff l.) a nd c ha ng e of c on ce nt ra tio n d ur in g p ro ce ss in g (R ed .% ) ( I− III , s up p le m en ta ry m at er ia l ( S up - p l.) ) P er .II I a nd IV : t he s tu d y p er io d s III a nd IV o f p ap er II I. O th er s ym b ol s as in T ab le s 13 a nd 1 4. M an ur e pH T S T O C T C IC N to t N H 4+ -N P t ot P O 4- P C a2 + K + M g2 + N a+ Pa pe r (% ) (m g/ L) Fe ed Sw in e 7. 1- 7. 4 0. 55 10 83 26 26 15 43 15 09 13 67 82 21 16 0 96 4 49 36 1 Ef fl. I 8. 3- 8. 6 0. 49 95 9 22 88 13 29 14 87 13 39 51 18 13 6 97 0 24 36 0 Re d. % 10 .5 11 .4 12 .9 13 .8 1. 5 2. 1 37 .8 13 .1 15 .2 -0 .6 50 .4 0. 2 Fe ed Sw in e 7. 9 1. 01 14 71 27 74 13 02 18 16 14 94 25 2 50 20 4 95 2 82 28 4 Ef fl. II 8. 8 0. 70 10 3 8 19 95 95 7 15 48 12 92 16 6 40 14 8 94 2 53 27 2 Re d. % H RT =9 d 30 .7 29 .4 28 .1 26 .5 14 .8 13 .5 34 .1 20 .0 27 .5 1. 1 35 .4 4. 2 Fe ed Sw in e 8. 1 0. 60 14 71 27 74 13 02 15 45 13 20 96 42 16 1 10 34 35 31 6 Ef fl. Su pp l. 8. 8 0. 51 10 89 20 70 98 1 13 06 11 50 64 5 34 10 3 10 50 14 5 32 1 Re d. % H RT =2 .3 d 16 .1 26 .0 25 .4 24 .6 15 .5 12 .9 32 .9 18 .2 36 .1 -1 .5 58 .2 -1 .8 Fe ed D ai ry 7. 6 2. 32 93 22 11 02 3 17 01 21 69 12 96 31 0 15 5 81 9 28 01 35 3 51 6 Ef fl. II I/ Pe r.I II 8. 5 1. 84 49 72 61 69 11 97 10 47 43 8 16 6 57 46 4 29 19 39 2 56 1 Re d. % 20 .7 46 .7 44 .0 29 .6 51 .7 66 .2 46 .6 63 .0 43 .4 -4 .2 -1 1. 2 -8 .7 Fe ed D ai ry 7. 9 1. 90 53 35 64 96 11 61 15 90 85 9 23 0 14 2 44 2 21 08 21 8 19 7 Ef fl. II I/ Pe r.I V 8. 8 1. 52 22 93 33 06 10 13 11 67 51 0 89 35 14 1 26 50 26 8 31 3 Re d. % 20 .0 57 .0 49 .1 12 .8 26 .6 40 .7 61 .4 75 .5 68 .0 -2 5. 7 -2 2. 6 -5 9. 0 MTT SCIENCE 29 55 4.2.3 Effect of treatment on manure nutrient content The nitrogen content changes of slurry manure occurred mainly due to ammonia volatilization during treatments. However, in the first study with swine manure there were almost no changes in nitrogen con- centrations during treatment (Table 17; I: Fig. 3A). In other studies with swine slur- ry manure, an average of 15% total N re- duction and 13% ammonium N reduction was observed due to treatment (Table 17, II, and supplemental data). The used am- monia stripping influenced N concentra- tions with dairy slurry manure treatment (III). Slurry NH4+-N content decreased be- low half its original concentration in the first treatment tank during the third study period, and the corresponding decline dur- ing the fourth period was 40% (Table 17; III: Table 3). This was due to the feedback dilution effect. During the course of the treatment (between tanks 1 and 6) neither the Ntot nor NH4+-N content of slurry ma- nure decreased significantly (III). Overall, effluent from the last tank contained on average 27 and 52% less Ntot and 41 and 66% less NH4+-N compared to the slur- ry manure before aeration in the two pe- riods studied (Table 17). Higher N reduc- tions were obtained with originally higher N content slurry (III, period III). Nitro- gen mineralization was observed in the case of slurry with a higher solids content (III, period III). Irrespective of the study, the amount of nitrate was low, with concentrations vary- ing between 9.9 and 31.2 mg/L in the first study with swine slurry manure (I) and less than 9 mg/L during dairy slurry ma- nure treatment (III). Accordingly, nitrate formation was low with the aeration rates used. However, the very low solid content of the slurry manure appeared to expose the material to easier nitrate formation (I). Of the Ptot 33–61% and of the phos- phate phosphorus 13–76% was removed as a result of biological treatment (Table 17). With swine slurry manure feeding, phosphorus contents were low due to the pre-treatment. More than 80% of the to- tal P of swine manure and nearly 90% of the phosphate phosphorus had been re- moved before biological treatment (study I). Despite the low phosphorus content of the feed, up to 38% of the swine ma- nure total P and 20% of the phosphate phosphorus was removed during biologi- cal treatment (Table 17). With dairy slurry manure, 47–61% of the total phosphorus and 63–76% of the phosphate phosphorus were removed, respectively, during the pro- cess (Table 17, III). The greatest Ptot con- centration changes occurred in the first tank irrespectively of the study. As was ob- served previously with respect to TS and carbon reductions, it appeared that phos- phorus reduction also depended on the in- itial content: the higher the concentration in the feed, the higher was the reduction. Magnesium and calcium ion concentration changes were similar to those of Ptot (Ta- ble 17). Predictably, monovalent Na+ and K+ ion concentrations during processing remained almost unchanged irrespective- ly of the study (Table 17). 4.2.4 Total bacterial counts and enteric indicator organisms in slurry Total bacterial counts Total bacterial numbers were enumerat- ed in dairy slurry manure. Total aerobic bacterial counts in raw dairy slurry varied from 2.95 105 to 3.24 106 cfu/g. Mean aerobic counts in aerated tanks varied from 1.88 105 to 6.17 106 (Fig. 11A; IV: Fig. 2). Total counts were increased (on Day 33 and 61 samples) or only slightly decreased (on Day 133 sample) compared to raw slur- ry and there were only small differences between the aerated tanks. On Day 172 total counts of samples with a short HRT value (6.7 d) were significantly decreased in the last treatment tank (Fig. 11A; IV). Enteric indicator organisms Swine slurry manure contained initial- ly higher viable counts of enteric bacteria compared to dairy slurry manure irrespec- 56 MTT SCIENCE 29 tively of time points of the study periods, with the exception of Enterobacteria (Fig. 11; IV: Fig. 2). Furthermore the reductions obtained (%) in the numbers of hygien- ic indicators due to treatment were high- er with swine slurry manure compared to dairy slurry manure, with the exception of fecal coliforms (Table 18; IV: Table 2). This could be exclusively due to higher original counts in swine slurry, since af- ter the treatment the numbers of indica- tor bacteria were higher in swine slurry manure than in dairy slurry manure (Fig. 11). Of the individual indicators, the num- bers of fecal streptococci remained high- er throughout the treatment compared to E. coli and other coliform bacteria in dairy slurry manure, although the opposite was observed in swine manure. Aeration treatment was conducted with fixed HRT values with swine slurry ma- nure, whereas with dairy slurry manure the HRT differed with one sampling time from the other samplings. The numbers of E.coli (attained reduction-% varied from 93.9 to 98.4) and total coliforms with swine slurry manure were significantly re- duced irrespectively of sampling time (Ta- ble 18), whereas with dairy slurry manure different HRT values influenced the result obtained: with higher HRT values (days 33, 61 and 130, HRT=43–48 days) the numbers of E. coli and total coliforms were significantly reduced, whereas with shorter HRT value (day 172, HRT=6.7 days) wide variation was observed in the numbers of these indicator microbes from one tank to the next and there was no decrease in the last tank compared to the original count (Fig. 11, IV: Fig. 2 and Table 2). Fecal coliform reduction in the last tank was 93.9% of the original count even with 1 100 10 000 1 000 000 100 000 000 0 1 2 3 4 5 6 0 1 2 3 4 5 6 0 1 2 3 4 5 6 0 1 2 3 4 5 6 Day 33 Day 61 Day 130 Day 172 cf u/ g E.coli Fecal coliforms Coliforms Fecal Streptococci Sulfite reducing clostridia Total counts 1 100 10 000 1 000 000 100 000 000 0 1 2 3 4 5 6 7 8 0 1 2 3 4 5 6 7 8 0 1 2 3 4 5 6 7 8 Day 58 Day 67 Day 71 cf u/ g A B Figure 11. The geometric means of fecal indicator bacteria and total counts (cfu/g) during aerobic treat- ment at different sampling times in untreated manure (0) and A. tanks (1−6) of the treatment system of dairy manure (HTR 43−48 days for sampling days 33, 61 and 130, 6.7 days for day 172=6.7, IV), and B. tanks (1–8) of the serial treatment system of swine manure (HTR 6.3 days, supplementary data). MTT SCIENCE 29 57 the shorter HRT (IV: Table 2) with dairy slurry manure, whereas with swine slur- ry manure more variation was observed in the reduction percentages (59.9–98.5%) (Table 18). Short retention time did not appear to in- fluence the reduction of fecal streptococ- ci with dairy slurry manure either. Their count was markedly reduced during treat- ment, except on day 61, when there was a marked variation in the numbers of fe- cal streptococci between the treatment tanks (Fig. 11). With swine slurry ma- nure good reduction percentages were at- tained in the numbers of fecal streptococ- ci (81.7–95.7%) without variation between the sampling dates (Table 18). Of all the indicator microbes, treatment had no or only little influence on sulfite- reducing clostridia, irrespectively of sam- pling time or manure type (Fig. 11 and Ta- ble 18, III: Fig. 2, III: Table 2). 4.2.5 Bacterial diversity dynamics Microbial diversity dynamics in separated dairy slurry manure and in aerated dairy slurry manure in different treatment tanks (Tanks 2 and 6) of the serial system were studied. The aim was to study changes in microbial community composition due to aeration and between the treatment tanks. In addition, the soil sample (the origin of an inoculant) microbial composition was studied and was compared to the micro- bial composition of the treatment tanks. Table 18. Statistical differences in numbers of enteric organisms between the different time points of the study period (Fig. 11B). Numbers of hygienic indicators (E. coli, Faecal coliforms, Coliforms, Enterococ- ci, sulfite-reducing clostridia and total counts) were analyzed with the MIXED procedure of SAS. Line- ar regression analyses (intercept, factor, R2 and p-values shown). Swine manure (supplementary data). Analysis of variance Linear regression Reduction Microbe Exp Day Level average (cfu/g) Intercept Factor R2 p (%) tanks 0−8 E. coli 58 4.09*102 a 9.69*102 -1.40*102 0.740 <0.003 97.9 66 8.36*102 a 2.18*103 -3.37*102 0.481 <0.038 98.4 71 4.71*102 a 1.05*103 -1.45*102 0.703 <0.003 93.9 Faecal 58 1.77*103 a 2.03*103 -64.7 0.072 <0.485 59.9 coliforms 66 3.59*103 a 1.03*104 -1.68*103 0.344 <0.097 98.5 71 1.41*103 a 2.74*103 -3.32*102 0.446 <0.049 85.9 Total 58 2.18*103 a 3.00*103 -2.05*102 0.430 <0.055 76.3 coliforms 66 4.42*103 a 1.25*104 -2.02*103 0.363 <0.086 98.5 71 1.89*103 a 3.79*103 -4.77*102 0.530 <0.026 87.9 Entero 58 1.77*102 a 3.34*102 -39.2 0.530 <0.026 81.7 bacteria 66 3.46*102 a 9.11*102 -1.41*102 0.424 <0.057 99.3 71 2.44*102 a 5.50*102 -76.6 0.819 <0.001 95.7 Sulfite 58 3.43*104 a 4.31*104 -2.27*103 0.082 <0.454 19.3 reducing 66 3.61*104 a 5.73*104 -5.28*103 0.781 <0.016 85.0 clostridia 71 1.48*104 b 1.96*104 -1.19*102 0.067 <0.501 26.0 Level averages between experiment days within a column having different letters indicate a significant difference at p< 0.05. † nd, not defined. 58 MTT SCIENCE 29 Changes in microbial community composi- tion in the 2nd and 6th treatment tanks The major influence of aeration was the decrease in the number of Firmictus in aerated slurry tanks compared to raw ma- nure before aeration and the increase in the numbers of Deinococcus-Thermus and Proteobacteria in aerated tanks (IV). Due to the aeration treatment, the num- ber of clones belonging to the phylum Fir- micutes decreased from 44% in untreated slurry to 6% in the 2nd tank of the serial system and continued to fall to 3% in the 6th treatment tank (IV: Fig. 3; IV: Table 3). The proportion of Deinococcus-Ther- mus increased from 10% in untreated slur- ry to 39% in the 2nd treatment tank, but fell to 29% in the 6th treatment tank. Cor- respondingly, the proportion of Proteobac- teria increased from 10% to 45% in the 2nd treatment tank and continued to increase up to 63% in the 6th treatment tank (IV). The species diversity of the Proteobacte- ria was also comprehensive. Clone librar- ies from tanks 2 and 6 included classes of Alpha-, Beta-, Delta-, Epsilon- and Gam- maproteobacteria (IV: Table 3; IV: Fig. 4). Deinococcus-Thermus Predominant OTUs detected in the 2nd and 6th tanks made up a single cluster in the phylogenet- ic tree, suggesting that the various Deino- coccus-Thermus representatives were close- ly related, irrespectively of their origin (untreated, tank 2, tank 6) (IV: Fig. 4). The proportion of the class Bacilli, mem- ber of the phylum Firmicutes, increased in the 2nd tank, but no Bacilli were present in the last treatment tank. Clones belong- ing to the phyla Bacteroidetes and Chlor- oflexi appeared in the 2nd treatment tank and Planctomycetes in the 6th treatment tank. Untreated slurry did not contain measurable levels of these phyla, where- as the soil from which the inoculum orig- inated contained Planctomycetes. How- ever, the Planctomycetes in the 6th tank and in the soil belonged to different class- es and families. Soil - the origin of the inocolumn The original inoculant was prepared from a soil-water suspension (III). Acidobacte- ria were the most abundant phylum of soil bacteria (38%), dominated by subgroup Gp6 (IV: Fig. 2; IV: Table 3). Acidobac- teria were followed by the phyla of Pro- teobacteria (29%), Gemmatimonadetes (15%), Verrumicrobia (6%), Planctomycet- es (3%), Firmicutes (3%) and Chloroplast (3%). The clone libraries of soil and treat- ment tanks did not share any sequence at the genus level, but some at the family level and several at the order level (IV). 4.3 Buffer system in the slurry and ammonia stripping 4.3.1 Ammonia stripping Ammonia stripping with the ammonia reference solution The constructed ammonia stripping de- vice was tested by using different refer- ence solutions (NH4+-N concentration from 300 to 1700 mg/L, II). The strip- ping performance with the used operation parameters was good (II). Over 90% am- monia removal percentages were reached irrespectively of the NH3 content of the reference solution (II: Fig. 2). However, the NH3 removal efficiency was shown to be directly dependent on the air and liquid temperatures (II). Ammonia stripping of the biologically treat- ed manure without chemicals Over 30% total ammoniacal nitrogen (TAN) removal by air stripping was shown to be possible if the pH of the biologically treated swine manure was above 8.9 (II: Fig. 3A). The NH3 remov- al with biologically treated swine slurry varied between 65 and 67% and TAN removal between 20 and 32% when the pH of the biologically treated manure was above 8.9 (II: Fig. 3A). With low- er pH values (pH-value variation from 8.6 to 8.9), 6 to 26% TAN removal was reached (II: Fig. 3A). MTT SCIENCE 29 59 However, with biologically treated dairy slurry (pH value 8.8 before stripping), only 6.2% of TAN was removed by air stripping (III: Table 3, III: Fig. 6B). The low removal percentage with dairy slur- ry was probably due to low ammonium concentration in slurry before stripping and also due to the high TS content of the slurry (III). Ammonia stripping of the biologically treat- ed manure with the chemicals used With biologically treated dairy slurry, MgO was added before stripping. This increased the slurry pH from 8.5 to 8.9 and facilitated 15.6% TAN removal (III: Table 3, III: Fig. 5B). Compared to the dairy slurry stripping result without MgO addition, the achieved result was better, but was still very low compared to the results achieved with biologically treat- ed swine slurry without chemical addi- tions (II). Repeated stripping procedure after first stripping Due to a decrease of pH after the first stripping, stripped swine manure was treated with 0.5 kg m-3 magnesium ox- ide (MgO) and stripped again. The use of MgO resulted in pH increase even to 9.2 (II: Table 2) before the second strip- ping. However, the pH reached varied widely between experiments. The sec- ond stripping decreased the slurry TAN by 15 to 26% after MgO treatment (II: Fig. 4A). The second stripping was followed by calcium hydroxide (Ca(OH)2) treatment to increase the pH of the twice stripped manure and then continued with a third stripping. In the third stripping, 29 to 58% of the TAN was removed compared with the TAN content before Ca(OH)2 addition (II, Table 2). Overall, after three consecutive stripping occasions, 59 to 86% of the original TAN of swine slurry had been removed (II: Table 2, Fig. 4A). 60 MTT SCIENCE 29 4.3.2 Buffer system composition in the slurry Buffer capacity curves were drawn in or- der to identify the acid and base species making a major contribution to the buff- er capacity in slurry (II). All the sam- ples showed differences in buffer capaci- ty when different amounts of TAN were present in a sample (Figure 12). Taking into account the acid dissociation con- stant (pKa) values (6.38, 10.32 and 9.25) for CO2/HCO3-, HCO3-/ CO32- and NH4+/NH3 , respectively, it appears like- ly that the peak at pH 6 to 7 was attribut- ed to HCO3-, the peak at 9 to 10 includ- ed the buffer capacity contribution from NH4+/NH3 and at pH 10 to 10.5 the peak was due to CO32- (Fig. 12; II: Fig. 5B). It was shown that the reduction in buffer capacity of the slurry was due to ammo- nia and carbonate removal during strip- ping (Fig. 12; II: Fig. 5A and B). Figure 12. Buffer capacity curves for untreated swine slurry (RL), biologically treated swine slurry (Pros), and the treated swine slurry after the first stripping (1.Str), second stripping (2.Str), and third stripping (3.Str) (II). * Slurry samples with carbonates previously removed. Pka values marked with vertical lines for CO2/HCO3-, HCO3-/ CO32- and NH4+/NH3 . 0 25 50 75 100 0 2 4 6 8 10 12 Bu ffe r c ap ac ity (m eq /L /p H ) pH RL 1.Str 2.Str 3.Str Pros RL* Pros* 1.Str.* 2 Str.* 3.Str.* pKa CO2/ HCO3-- 6.38 pKa NH4 +/NH3 9.25 pKa HCO3-/CO3 2- 10.32 MTT SCIENCE 29 61 5 Discussion 5.1 Solids separation before aeration treatment As a pre-treatment, solids separation of slurry manure was carried out in order to achieve suitable solids content prior to bi- ological processing (I and III). The goal was to reduce manure TS content to 1% or lower in order to achieve efficient bio- logical treatment. According to Zhu et al. (2008), in order to reduce aeration time and rates the manure solids content should not exceed 1%. Oxygen transfer rate and efficiency are affected by factors such as the surface area in contact with air, mix- ing, temperature, and the amount of sol- ids or other constituents in the manure (Martin and Loehr, 1976; Westerman and Zhang, 1997; Ndegwa, 2004). According to Zhu et al. (2005), the oxygen trans- fer coefficient in slurry at 4.0% total sol- ids level was reduced to about a quarter of the value in slurry at 0.5% total solids lev- el. According to those authors, the energy consumption for slurry with 4% total sol- ids is nearly four times that for slurry with 0.5% solids content in order to achive the same DO level in the treated slurry. Finer solid particles also decompose faster and to a greater extent than coarse particles (Westerman and Zhang, 1997). Accord- ing to Ndegwa (2004), with solids separa- tion improved aeration efficiency and bet- ter contact between the microbial biomass and the dissolved substrate was achieved, thus boosting the aeration treatment and biostabilization. Sedimentation alone decreased the swine slurry manure dry matter content to ca. 1% or below (I). The achieved good sedi- mentation result was probably due to raw slurry initially containing a low solids con- tent (1.85%) before the sedimentation step (I). This agrees with the results of Ndegwa et al. (2001), who found that the best re- sults in manure separation were achieved with 1.0 and 2.0% solids levels. Natural settling reduced swine manure TS con- tent by 70%, total P content by 80% and almost 90% of the phosphate P (I). The results are in agreement with the previous results of Ndegwa et al. (2001) and Zhu et al. (2004). Dairy manure mechanical separation with a screw press resulted in an average of 2.9% TS content in the liquid phase (III). Separation removed 33% of the ma- nure TS and 12.5% of the Ptot. The ob- tained results were in agreement with pre- vious research conducted by Møller et al. (2002) with freshly collected dairy ma- nure. Screw press separation efficiency is dependent on its ability to retain particles of certain size. According to Møller et al. (2002), the screw press separator only re- tained particles > 1 mm, which were low in phosphate compared to smaller parti- cles. This was also probably the explana- tion for the achieved low nitrogen separa- tion result (less than 5% of Ntot separated) with a screw press (III). It was likely that after the solids separation step, finer compounds remained in the liq- uid fraction. Those compounds probably included odor compounds and compounds that were precursors for odor generation, as was reported e.g. by Zhang and Westman (1997) and Jacobson et al. (2001b). Sup- porting this, mechanical separation meth- ods have been considered to have only a limited capacity for separating out odor- ous compounds (Zhu et al., 2001; Nde- gwa et al., 2002). 62 MTT SCIENCE 29 5.2 Theoretical basis for the designed reactor type The reactor type used in the present study was a system of six continuously fed aerat- ed tank bioreactors grouped by serial con- nection conducted with feedback (I, III). This reactor type was designed on the ba- sis of data in the literature on reactor per- formance. The goal with the reactor design was to achieve stable and efficient treat- ment in treatment facilities of comparative- ly small size. 5.2.1 Several bioreactors grouped by serial connection Chemical reaction theory states that any monotonic reaction proceeds most rapid- ly in a plug flow reactor (PFR) due to the substrate gradient developed with that con- figuration. Therefore, an ideal PFR con- figuration would achieve specific perfor- mance goals in a smaller reactor volume than any other configuration (Levenspiel, 1972). However, Erickson and Fan (1968) recommended the use of a tanks-in-series configuration for activated sludge systems to reduce back mixing and approximate the plug flow-type substrate gradient. Hydraulic performance of a CSTR reactor can be im- proved by increasing the number of CSTRs in series (Reynolds and Richards, 1996). For the same total reactor volume (Vt) it is pos- sible to approach the performance of a PFR (in terms of retention time) by increasing the number of CSTRs (n) in series (in this case each CSTR in the series has a volume of Vt/n) (Reynolds and Richards, 1996). CSTRs in series have commonly been used in biological treatment of industrial waste- water, such as activated sludge basins which are cascade connected (Abu-Reesh, 2010). This arrangement of reactors offers a num- ber of advantages such as increased stabili- ty of the treatment plant when subjected to pulse load and also enhanced degree of deg- radation by adoption of activated sludge re- cycling (Abu-Reesh, 2010). Based on the above literature, a configu- ration of six tanks in series was designed. This was assumed to provide small reactor volume of the plug flow-type reactor de- sign, but more stability compared to plug flow design. 5.2.2 Feedback In the present study feedback was intro- duced in the system in order to increase the stability of the treatment process. In biological systems it is well known that most parameters must stay under control within a narrow range around a certain optimal level under certain environmen- tal conditions (Thomas and D’Ari, 1990). Deviation from the optimal value of the controlled parameter can result from the changes in internal and external environ- ments. Homeostasis in biological systems maintains thermal, chemical, and biolog- ical conditions through feedback (Thom- as and D’Ari, 1990). In practice, feedback was carried out by recycling a part of the treated material back to the process. In this respect, the designed reactor system rather resembled a recycle reactors design (Hill and Root, 2014) than an active sludge system (Reyn- olds and Richards, 1996). 5.2.3 Use of limited aeration Limited aeration was applied in the de- signed series reactor system. Manure aer- ation using limited aeration is used for manure stabilization (Ndegwa et al., 2007), odor reduction (Ndegwa, 2003), and hygienization purposes (Heinonen- Tanski et al., 2006), most frequently car- ried out, however, in single tank systems with continual or intermittent aeration. More efficient biological treatment tech- nologies are based on nitrification-den- itrification processes, in which nitrogen is lost, not captured (Reynolds and Rich- ards, 1996). This study aimed at deter- mining the level of aeration that would accelerate humification reactions, but would not lead to nitrate formation. Re- duced odor and improved hygiene were also targeted goals. MTT SCIENCE 29 63 5.3 Benefits of the serial treatment system The functional goals of the designed serial manure treatment system were to achieve A) treatment stability towards system fail- ures, B) treatment efficiency (short HRT), and C) improved process control. Besides the functional goals, the aims of the bio- logical treatment were to achieve: 1. ef- fective odor reduction; 2. good hygienic quality; 3. stable treatment product con- currently, and 4. preservation of nitrogen in NH3 /NH4+ form in a liquid phase in order to continue the treatment with am- monia stripping. 5.3.1 Treatment stability towards system failures It was clearly shown on the basis of the obtained ORP-values that the six tanks system in series provided stability for the treatment of dairy manure (III). Com- pared to the single tank system, in a seri- al system changes in the subsequent tanks occurred with a delay, thus indicating the dynamic character of the treatment system (III). Partly this was achieved with serial configuration, which effectively prevented material from being mixed between tanks and allowed only part of the material to flow from one tank to the next in a given time (III). Feedback also functioned evi- dently as a buffer mechanism against pro- cess interferences and was able to provide a system very resilient towards external in- fluences, as described by Åström and Mur- ray (2009). In activated sludge treatment systems, a part of the settled material, the sludge (of- ten called return activated sludge), is re- turned to the head of the aeration system to re-seed the new wastewater entering the tank (Reynolds and Richards, 1996). In the present system, feedback was not only the sedimented microorgnisms (activat- ed sludge), but part of the effluent (par- tial recycle stream) exiting from the last tank of the serial system (I, III). In chem- ical reactor design, however, partial recy- cle stream is used when a substrate cannot be completely processed in a single pass, such as with an insoluble substrate (Hill and Root, 2014). These reactors continue to move the same substrate through the re- actor so that the effective contact time is high enough to allow the substrate to be processed. Recycle reactors also allow the reactor to operate at high fluid velocities (Hill and Root, 2014). This is important because it minimizes the bulk mass trans- fer resistance to the transport of the sub- strate (Conroy, 1997). 5.3.2 Treatment efficiency The HRT time of the designed system was measured in days (I, III), indicating efficient manure treatment. Factors that might influence the treatment efficiency of a serial tank system with recycle were feedback dilution effect (III) and the indi- cation of different decomposability of or- ganic matter in different treatment tanks based on the obtained redox profile (III), and the community compositional chang- es observed in different treatment tanks (IV). The feedback significantly reduced the concentrations of carbon and cationic ions in the first treatment tank (III, I). Thus, it was probable that feedback enhanced carbon reduction due to a dilution effect in a system with high organic load (as ex- plained in III). If ammonia is a process in- hibiting factor, its removal from the recy- cle stream leads to a more efficient process. The feedback TAN concentration was de- liberately lowered by NH3 stripping in or- der to improve the treatment process and to avoid NH3 inhibition (III). Ammonia inhibition has been commonly acknowl- edged as an inhibition factor in anaerobic digestion, as methanogens have been re- ported to be sensitive to ammonia (Chen et al., 2008). However, studies concern- ing ammonia inhibition in aerated live- stock slurry are scarce, although ammonia in aerated slurry with higher pH is evi- dently present to a higher degree in am- 64 MTT SCIENCE 29 monia form compared to anaerobic diges- tion. Free ammonia has been suggested to be the main cause of process inhibition, since it is freely membrane-permeable and may diffuse passively into the cell, caus- ing proton imbalance and/or potassium deficiency (Sprott and Patel, 1986; Gal- lert et al., 1998). According to Kokkonen et al. (2006), accumulation of ammoni- um in solution affected the mobilization rate and final reaction equilibrium. They showed that accumulation of the gaseous reaction products, e.g. NH3 -N and CO2, in the system, instead of volatilization to surrounding air, slowed the reaction rate down. Indication of the different decomposition in different treatment tanks suggested that easily degradable material diminished al- ready in the first treatment tanks, where- as more refractory material accumulated in the last tanks (III). Composting studies have revealed organic matter decomposi- tion and microbial succession (e.g. Nakasa- ki et al., 2005). The microorganisms con- tributing to organic matter decomposition have changed as the composting progress proceeds (Nakasaki et al., 2005). The ob- served community compositional chang- es between different treatment tanks (IV) support the idea that microbial succes- sion was spread within the different treat- ment tanks of the serial system. Whether or not this would lead to decreased micro- bial competition between successive tanks, and thereby increase treatment efficiency, remained however uncertain. 5.3.3 Odor reduction It was shown that biological treatment in a series of continuously fed aerated tank re- actors efficiently reduced the odor of the swine slurry manure (I). Concerning the odor reduction, serial configuration with feedback was one probable reason for in- creased treatment efficiency. Although the obtained results were preliminary and did not extend to individual compounds, it was noticeable that odor was reduced from one tank to the next, clearly waning to- wards the end of the process (I). On a the- oretical basis the following was thought to provide an explanation: Based on odor characteristics even tiny amounts of odor- ants have been found to produce a strong smell because the additive and/or synergis- tic effects of hundreds of compounds in- crease the strong odor intensity (Schiffman et al., 2001). Based on the flow characteris- tics of serial tank reactors, only part of the solution flows from one tank to the next at a time, and consequently the mixing of the entire volume is prevented (I, III). Therefore, an example can be presented in which odor compounds concentrations are reduced according to a decreasing geomet- ric series. In the first tank odorants are di- luted by half due to the feedback dilution effect (assuming feedback only contains stabilized organic matter and its volume equals that of the feed volume). Only part (for example one tenth) of the odorants ex- isting in the 1st tank flows to the 2nd tank. From the second tank again only 1 tenth of the odorants flow to the next container, and so on. Thus, in the last (6th ) tank, the residual concentration of odorants is only 0.5*10-6. Thus, it is reduced to one part in a million of the original concentration. 5.3.4 Advanced process control With regard to process control the serial system provides several benefits over a sin- gle tank system (III). The obtained results showed how process control became easier with a serial system compared to a single tank system. It was concluded that process failure was easily predictable (III) and the dynamic system gave time to start correc- tive measures in time. Moreover, the seri- al treatment system allows advanced pro- cess control (over aeration) according to demand. Aeration demand in the first tank was higher than in the following tanks, and thus aeration could be adjusted to a higher level in the first tank and to low- er levels in the last tanks according to de- mand, thereby saving energy (III). Howev- er, drawbacks were reported (Yoon, 2011) MTT SCIENCE 29 65 relating to the high aeration demand in the first tank: 1) due to bubble coalescence, the high upstream oxygen demand makes it hard to dissolve enough oxygen to sup- port biological COD degradation, 2) due to the bubble coalescence and high sur- factant-like molecules in wastewater, ox- ygen transfer efficiency is low upstream, 3) the high ratio of nutrients to microor- ganisms causes higher diffuser fouling in upstream, which again makes it harder to dissolve enough oxygen in the local space. 5.4 Process conditions – Biological and physico- chemical factors in a serial system 5.4.1 Treatment start-up and inoculation The slurry manure was treated biological- ly in the serial tank system. Before start- ing the treatment process, the tanks were filled with microbial seeding material (I). Manure aeration is commonly conducted without using any external microbial ad- ditives or seeding (e.g. Zhu et al., 2005). Microbiological additives have been used for odor reduction purposes (Ritter, 1981; Zhu et al., 1997; Zhu, 2000). However, it is acknowledged that commercial diges- tive deodorants which contain bacteria and enzymes added in manure have had limit- ed success in controlling odor (e.g. Ritter, 1981). It has been reported, neverthless, that a commercial microbiological addi- tive has functioned effectively during the manure composting process (e.g. Nakai et al., 2004) and even very small amounts of additives have resulted in a good result (Wang et al., 2011). The use of seeding is a common procedure e.g. with anaerobic digestion and activat- ed sludge during the start-up phase to en- hance degradation activities and to shorten the start-up times and treatment periods (Gerardi, 2003a; Pijuan et al., 2011). In the present study, the aim was to carry out seeding in order to increase the degrada- tion potential in treatment reactors (IV). The idea was in particular to improve the odor reduction potential of the biological treatment. Moreover, slurry manure itself (anaerobic source) was not seen as the best source for enriching the most suitable mi- crobial population for aerobic treatment. Seeding material was produced from an agricultural soil with a long history of ma- nure applications and an abundant earth- worm population (I, IV). The aim was to take advantage of the large reservoirs of genetic information in soil (Torsvik and Øvreås, 2002). It is probable that soil (es- pecially a soil that has been under inten- sive agricultural use and received abundant manure applications) contains capacity for utilization of diverse odor compounds. Preparing the seeding material A seeding material was produced by us- ing soil as a starting material (soil:water ratio 1% v/v). The procedure is de- scribed in more detail in publication I. Naidu at al. (2010) described a method- ology to produce aerated tea compost, a water-based compost extract containing a high population of beneficial microbes. The compost:water ratio was 1:5 (w/v). Compared to the tea compost production (Naidu et al., 2010), the used soil:water re- lation in the present study was much lower. The number of total bacteria in tea com- post was105 mL-1 (Naidu et al., 2010), but was not measured in the present study. Seeding and the treatment start-up In the present study, the start up-phase was to fill the treatment tanks with the pro- duced effluent. Thereafter, slurry manure feed was started with a continuous feeding rate (with small amounts of slurry, 20−50 L/day), monitoring the ORP and odor for- mation simultaneously and adjusting the feeding accordingly (I). Regarding anaer- obic digestion, a general guideline is that the seed material should be twice the vol- ume of the fresh manure slurry during the start-up phase, with a gradual decrease in amount added over a three-week period 66 MTT SCIENCE 29 (DaSilva, 1979). Thus, the used seeding volumes in the present study were much higher related to feed raw manure volumes (the feeding rate of 50 L/day represents one tenth of the first tank volume and one sixtieth of the entire system volume). Si- multaneously recycling stream (feedback) equivalent to the feeding volume was recy- cled from the last tank to the first tank (I). In the composting process, the desired ac- tivities have been achieved with even very small amounts of inoculant use with pre- selected microorganisms. According Wang et al. (2011), 57.5% faster lignocellulose degradation was reached by inoculating compost with a lignocellulolytic fungus. Fungal suspensions with a concentration of 1 × 109 colony-forming units (cfu) mL−1 were used at a concentration of 5 mL kg−1 of compost during the building of heaps. In another composting study, each com- posting heap was inoculated with a volume of microbial suspension to reach a level of 106 to 107 colony-forming units g−1 (cfu g−1) of waste (Vargas-García et al., 2006). In a study of Sasaki et al. (2006), a mixture comprising 400 kg of beef cattle manure and chaff, and 100 kg of mature beef ma- nure compost was well mixed and then mixed with 10 kg of microbial addition. The reason for successful inoculant use in composting (e.g. Wang et al., 2011) but unsuccessful use with manure (microbio- logical additives added in manure for odor reduction, e.g. Zhu, 2000) is probably due to material differences between these two materials and differences in treatment con- ditions. Not only does aerobic composting occur under aerobic conditions (and added inoculants are predominantly aerobes), but respiration intensity is also probably rela- tively low in dry compost material com- pared to stirred slurry manure. In fact, the treatment conditions are comparable to submerged (liquid) fermentation (liquid manure) and solid substrate fermentation (composting) (fermentation types reviewed e.g. Raimbault, 1998; Subramaniyam and Vimala, 2012). For this reason the seeding procedure in the present study was carried out as im- plemented (I) by adding only a small vol- ume of slurry manure to a large volume of seeding material and conducting feedback in order to increase the survival of aerobes in a liquid state with high organic load. In the present study with the start-up phase and with the seeding material prep- aration, the question was probably also one of adaptation (I). Several studies have re- vealed the importance of bacterial adapta- tion to a wide range of ammonia concen- trations (Angelidaki and Ahring, 1993; Angenent et al., 2002), but have not deter- mined whether the adaptation was a con- sequence of metabolic transition of the al- ready existing microbial population or of the growth of new cultures adapted to the different ammonia concentrations (Rajag- opal et al., 2013). Viable bacterial counts Total counts were enumerated in order to provide basic information on the num- bers of bacteria in slurry during treat- ment, as well as on overall microbial func- tion. In general there were higher total counts in aerated tanks than in raw slurry (IV). This was also supported by the 16S rRNA results (IV). With short HRT val- ue (HRT=6.7), however, total counts sig- nificantly decreased in the last treatment tanks of the serial system. Perhaps treat- ment time was sufficient to inactivate an- aerobic/ facultative anaerobic species in aerobic conditions in the last tanks of the serial system, but too short for aerobes to evolve. Indicator microorganisms were used as a marker of the hygenic state of the treat- ment and in order to provide evidence for the possible occurrence of ecologically sim- ilar pathogens in a system. Differences in enteric counts were observed between the MTT SCIENCE 29 67 slurry manure type, between the indica- tor organisms and between the different HRT values (IV). Higher numbers of en- teric bacteria in swine manure compared to dairy manure in the present study were probably related to the low solids content of the dairy slurry manure (found e.g. Jones, 1982). The levels of E. coli, fecal (thermotolerant) coliforms and total coli- forms in raw dairy slurry were considera- bly lower in the present study than in the earlier study of Heinonen-Tanski (1999). Comparing individual indicators, fecal streptococci have been more resistant to stress than E. coli and the other coliform bacteria (OECD/WHO, 2003), and this was also seen in the results with dairy ma- nure (Fig. 11A). With swine manure, how- ever, the opposite result was obtained (Fig. 11B). Of all the indicator microbes, treatment had no or only little effect on sulfite-re- ducing clostridia, irrespectively of sam- pling time or slurry type (Fig. 11, Table 18). Sulfite-reducing clostridia are spore- forming and therefore have been more re- sistant to treatment and disinfection pro- cesses (Gould and Hurst, 1969; NRC, 2004). Bacterial community composition A shift in the bacterial community com- position from obligate or facultative anaer- obes to aerobically growing bacteria was observed due to aeration (IV). Differenc- es in community composition between the aerated tanks were also observed (IV), al- though the result is preliminary and needs confirmation. It has previously been re- ported that microbial communities have been able to change very rapidly, and that variability in community composition has changed even more rapidly (e.g. Redford and Fierer, 2009). Research areas under in- tensive study at the moment (such as the colonization process of the infant gut mi- crobiome) (e.g. Koenig et al., 2011) could provide new interesting insights on this subject area in the future. How microbi- omes are established and maintained, is one of the leading questions in biology (Scheuring and Yu, 2012). 5.4.2 Aerobic decomposition under limited aeration and humification The process of decomposition is initially rapid, but slows down considerably as the supply of readily decomposable organic matter becomes exhausted (USDA/NRCS, 2000). Consequently, the last tank in the series contained further degraded, more stable material compared to the material in the first tank, and the aeration demand in the last tank in the series was obvious- ly less than in the first tanks (III). There- fore, it was also very likely that there were differences in compound composition be- tween the tanks. The observed community compositional changes between different treatment tanks supported this (IV). How- ever, further research is required to identi- fy the compound composition changes in the systems. Humification During the treatment process slurry ma- nure color changed from gray to dark brown, indicating humus formation (I). Humus formation is often considered to occur solely under anaerobic conditions (e.g. Rajeshwari and Balakrishnan, 2009). However, if aeration is kept within a lim- ited range, humus-like condensation prod- ucts may also form under aerobic condi- tions (Tölgyessy, 1993). In the present study, humification was studied using spectroscopic methods only (I). The observed dark color was linked to a recondensation process forming hu- mic substances such as fulvic and humic acids (I). The typical dark color of humic and fulvic acids is due to the presence of aromatic nuclei (Stevenson, 1984). Humic substances behave like weak acid polye- lectrolytes and the occurrence of anion- ic charged sites accounts for their ability to retain cations (Hayes and Swift, 1978). 68 MTT SCIENCE 29 Thus, this behavior probably enhanced coagulating behavior (complexation with divalent cations in a system), as observed during biological treatment (I). Problematic compounds in manure treat- ment are those low molecular weight com- pounds that are electrically neutral, high- ly water-soluble and difficult to remove by precipitation or filtration (Zhu et al., 2001; Ndegwa et al., 2002). These compounds include volatile fatty acids, low molecu- lar weight carbohydrates (simple sugars) and other dissolved compounds, e.g. or- ganic acids, phenols, nitrogen and sulfur compounds, low molecular weight pro- teins etc. that have been considered to be responsible for most offensive odor emis- sions (Zhang and Westerman, 1997; Jacob- son et al., 2001b). Aeration is used for odor control (I, III), because these compounds are easily consumed by microorganisms (Jacobson et al., 2001b). Microbial me- tabolites of these processes are probably involved in humification processes. Hu- mification is enhanced by high amounts of soluble organic compounds and limited aeration, and probably serves as a natural coagulant facilitating the precipitation of these weakly precipitating compounds (I). In soil conditions the humification pro- cess has been found to be generally very slow, taking years to reach completion (Tölgyessy, 1993). Under controlled con- ditions, as in a compost environment, hu- mification proceeds much faster, in weeks (e.g. Diaz et al., 2011). However, during slurry aeration, stabilization proceeds even faster, in days (Ndegwa et al., 2007a). The obtained results showed color intensifica- tion occurring in a very short time, sug- gesting a very rapid humification process (I). The reasons for accelerated humifica- tion processes during slurry aeration were probably the low C/N of the raw materi- al (C/N = 2), small particle size of the or- ganic material, sufficient O2 concentra- tion due to aeration (but not too intensive), and high water content. Moreover, aera- tion mixed the suspension thoroughly (I), which was probably the most important factor. ORP values under limited aeration The measured ORP values ranged be- tween -200 and +300 mV in treatment tanks when the dairy manure was treat- ed (III). ORP in separated slurry before aeration treatment was circa -200 mV and ranged from 0 to 200 mV in treat- ment tanks when the process functioned in the desired way (III). It was devised that ideally the redox-values of the treatment tanks should be kept as follows: 1st tank at zero or slightly above, 3rd tank at around 100−180, and the 6th tank at 180−200 mV (III). These ORP values were within the range reported by Ndegwa et al. (2007a), using limited aeration, and on a level high enough to reduce the odor potential in the manure (at +35 mV or higher) according to Zhu et al. (2002). The measured ORP values remained be- low +225 mV and no significant amount of nitrate was formed (I, III), although ORP values were on a level favoring nitri- fication (III, Fig. 4). Therefore nitrate for- mation was probably prevented by other factors. According to Strauss (2000), ni- trification rates were influenced by many factors: the template factors including pH, temperature, and dissolved oxygen and within these constraints C:N ratio, organic carbon availability, and nitrogen availability in the system. Many studies have revealed the negative effect of organ- ic carbon (in particular labile organic car- bon) on nitrification (Strauss and Dodds, 1997; Strauss, 2000). The mechanism re- sponsible for the inhibition of nitrification when organic carbon is abundant is prob- ably increased competition between nitri- fying and heterotrophic processes (Strauss, 2000). This was probably also the reason in the present study, although many mech- anisms were evidently involved in a com- plex system. The obtained characterization results revealed Nitrosomonas in the sec- MTT SCIENCE 29 69 ond treatment tank but no nitrifying or- ganisms were detected from the sixth treat- ment tank (IV). The low nitrate formation was an aim be- cause nitrate cannot be removed by ammo- nia stripping. Likewise, the nitrate which returns with recycle back to the first treat- ment tank is vulnerable to denitrification due to the low ORP value in the tank. 5.4.3 pH and its relationship with the release of NH3 and CO2 from the system In the present study, significant pH in- crease due to aeration was observed (I, II, III). The pH change in aerated manure slurry is controlled by a complex chemi- cal buffering system (Husted et al., 1991; Sommer and Husted, 1995; Paul and Beauchamp, 1989). Probably the most im- portant role in slurry is played by the am- monium bicarbonate system (II, Husted et al., 1991). Ammonia in solution is neutral- ized by dissolved CO2 to form ammoni- um bicarbonate, keeping the pH at around neutral. An increased CO2 release from the system in relation to NH3 leads to a pH increase, whereas the opposite leads to a pH decrease. This is explained accord- ing to the maximum solubility of these two components (II). The solubility dif- ference between these two components is 1.7 g CO2 L−1 compared with 535 g NH3 L−1 in pure water at 20°C at 1 atm pressure (Chang, 2000). In fact, the system is an interrelationship between the equilibrium constants, Hen- ry’s law, and the weak acid-base relation- ship. These factors apply both in biological aeration treatment as well as in ammonia stripping processes (explained in more de- tail in section 1.5.3, Physico-chemical fac- tors involved in treatment reactors). In aerated tanks, the solution pH affects the proportions of carbonate and ammo- niacal nitrogen which are present in bicar- bonate and ammonia form (Figs. 5 and 6). Above a pH-value of 8.5, the proportion of bicarbonate decreases with increasing pH (Fig. 6; Chang, 2000); at the same time the proportion of NH3 increases (Fig. 5; Chang, 2000). Thus, both CO2 and NH3 are volitilizated from the system. Howev- er, due to the solubility difference between these two compounds, more CO2 is voliti- lizated in relation to NH3 (Chang, 2000). 5.5 The role of manure buffer system in manure treatment In the present study, it was shown that bio- logical aeration treatment facilitated a slur- ry manure pH increase that in turn fa- cilitated ammonia removal, reducing the needed chemical consumption (II). Al- though the manure buffering system plays a central role controlling manure pH change and coagulation processes, and causes substantial chemical consumption (Husted et al., 1991; Sommer and Hus- ted, 1995; Paul and Beauchamp, 1989), the number of studies in which manure buff- ering capacity has intentionally been re- duced is limited. Buffer composition In the present study, it was shown that ma- nure buffer capacity was composed of TAN and the CO2, HCO3-, CO32− system (II). The obtained result was in agreement with the study of Husted et al. (1991). In their study, the major buffer components in ma- nure were found to be ammonium, bicar- bonate and a solid phase of carbonate. It was also shown that the reduction in buff- er capacity of the slurry was due to ammo- nia and carbonate removal during ammo- nia stripping (II). Selection of a precipitating agent In the present study, MgO was used to in- crease manure slurry pH and to precipitate the residual P in the effluent (II and III). MgO addition was performed either after biological treatment (III) or after the first stripping cycle and subsequent nitrogen re- moval (II) using a dose of 0.5 kg m-3. The 70 MTT SCIENCE 29 most important reason for the selection of a divalent rather than a trivalent precipitant was the plant availability of the formed pre- cipitate (Dao et al., 2001; Hyde and Mor- ris, 2004). Although divalent ions precip- itate less organic matter than the trivalent ions, they are reasonably effective P precip- itants. The other reason is that aluminium and iron salts, although commonly used for solids removal in waste water treatment, are rather ineffective in manure treatment because of the high buffer capacity in slur- ry manure. Due to high pH and a high- ly buffered system, too rapid polymeriza- tion occurs, creating insoluble precipitated aluminium and iron polymers, resulting in Al/Fe(III) being surrounded and there- by neutralized by negatively charged ox- ygen atoms/hydroxides, with consequent loss of flocculation/coagulation function (Jiang and Graham, 1998). Therefore ma- nure treatment with Al/Fe salts is in gener- al inefficient and requires rather high dos- es of Al/Fe. Influence of the buffer system on chemical use in ammonium separation The results showed that a better precipita- tion result was obtained with MgO dosing after first stripping (II) compared to MgO dosing after biological treatment (III) when using the same amount of MgO. This sup- ported the ammonia reduction effect on the need for chemicals. It appeared that ammonia stripping also affected the pH increase. After ammonia stripping, higher pH values were reached with MgO dosing compared to MgO dosing after biological treatment (II and III). 5.6 Sequential nitrogen separation Air stripping requires a high pH for the NH4+ to separate in the tower as NH3 gas. It is difficult to change the pH of the ma- nure because of the high buffering capacity of slurry manure (e.g. Sommer and Husted, 1995). Large amounts of chemicals are re- quired, which makes the process economi- cally unprofitable. As NH3 is removed, the pH of the solution decreases (also observed in II and III), and the effectiveness of the tower to separate N is diminished. In order to reduce the needed chemical consumption a new sequential stripping procedure was proposed. The sequential treatment scheme included (i) increasing the slurry pH without chemical use and with part of the slurry buffer capacity re- moved and (ii) increasing the slurry pH with chemical treatments in between the stripping cycles (II). It was shown that over 30% total ammoni- acal nitrogen removal by air stripping was possible without chemical use, if the pH of the biologically treated swine manure was above 8.9 (II). The slurry was further sub- jected to repeated cycles of stripping with MgO and Ca(OH)2 additions after the first and second strippings, respectively, to in- crease slurry pH in between the stripping cycles. After three consecutive stripping cy- cles, 59 to 86% of the original ammonium had been removed (II). It was important to observe that the buffer system in slurry manure was shown to con- sist of NH3 and carbonate systems and it was also shown that these compounds were removed during biological treatment and stripping cycles (II). Therefore, less chem- icals are needed when slurry manure pH is increased sequentially between the strip- ping cycles than when changing the slur- ry ammonium completely to NH3 before stripping. A patent was applied for this ob- servation (Kokkonen et al., 2013). In addition to ammonia removed by air stripping, part of the ammonia (on aver- age 15% total N and 13% ammonium N; Table 17, II and supplementary data) was released during aeration treatment. This ammonia can be preserved for further use, if the NH3 released during the biologi- cal treatment is collected and led to an air scrubber. MTT SCIENCE 29 71 5.7 Hygiene One of the main goals related to biolog- ical aeration treatment was to improve the hygienic quality of the slurry (IV). Hygienic quality measurements consist- ed of commonly applied methods based on enumeration of enteric indicator or- ganisms by cultivating samples on a spe- cific agar and defined temperature for a certain time period. Pathogen reduc- tion can be reached by various means, al- though the method was not directly ad- dressed towards pathogen reduction. The successive procedures have had a hygiene increasing effect (e.g Böhm, 2008). The hygiene influences of treatment steps and their ability to inactivate pathogens are evaluated below. Solids separation In the present study, hygiene indicator organisms were analyzed in dairy slur- ry after the solids separation step (IV). Solids separation probably improved hy- giene in the liquid fraction as indicat- ed by the lower levels of indicator organ- isms compared to the reported average counts in raw slurry (IV). However, in some studies pathogens have been found both in the solids and in the liquid frac- tions of the source separation systems (e.g. Letourneau et al., 2010), indicating that solids separation does not necessar- ily lower the indicator organism counts in the liquid fraction. The other mech- anisms that probably resulted in lower counts were related to storage time before slurry feeding in a slurry treatment sys- tem (IV). However, high solids content and low temperatures have been found to promote survival of pathogens (Strauch, 1991). According to Jones (1982), survival was greatest at temperatures below 10°C and in slurries containing more than 5% solids. Aeration treatment in a series of continous- ly fed aerated tank reactors The obtained results after biological aer- ation treatment revealed good hygien- ic state although the reduction percent- ages reached were not high, particularly in the case of dairy manure due to its relatively low initial enteric counts (IV, Fig. 11 and Table 18). This was in agree- ment with the results of Heinonen-Tan- ski (1999). Aeration treatment has been commonly been applied to achieve re- duction in the counts of intestinal bac- teria or viruses (e.g. Heinonen-Tanski et al., 2006; McGarvey et al., 2007). Path- ogen reduction during aeration was prob- ably due to a variety of factors such as O2 sensitivity, temperature, high concentra- tion of free ammonia, and high pH (e.g. Jenkins et al., 1998). Martens and Böhm (2009) concluded on the basis of data presented by Meyer (2001) that aeration treatment should be operated in at least two vessels connect- ed in series in order to achieve a suffi- cient exposure time and to avoid hydrau- lic short circuits during the addition and removal of slurry during operation. Based on this conclusion, one could expect a six treatment tank system connected in series to provide a means for achieving a good hygiene result. The present aerated treat- ment system also contributed to aspects such as pH shift, high redox potential and antagomism, which were listed by Martens and Böhm (2009) as factors ex- pected to lead to a more or less rapid in- activation of pathogens. Ammonia stripping The effluent solution after ammonia stripping was not evaluated based on hy- giene indicator measures. However, in- creased temperature has been found to be the most effective factor in pathogen inactivation (Martens Böhm, 2009). Al- though the residence time in ammonia stripping was relatively short, the repeat- ed chemical additions (pH shift) were agood hygienization verification. Howev- er, further research is needed on the effect of ammonia stripping on hygienization. 72 MTT SCIENCE 29 5.8 Applicability of the developed manure treatment technology The aim of this section is to evaluate the novel treatment system (I–IV) in a broad- er sense and to analyze its practical advan- tages. Benefits and weaknesses of the pro- cess steps are considered. Solids separation – effect on manure use as a fertilizer The obtained results demonstrated the difficulty to get all the material treated. It was evident that part of the solid ma- terial of both raw swine and dairy slur- ry was sedimented on the bottom of the manure storage tank and into a contain- er before the manure was pumped for- ward to the first treatment tank (I−III). Treating all the slurry requires efficient mixing in the storage and reservoir tanks from where it is pumped forward in the process. Mechanical separation of dairy manure did not affect the manure N content al- though it lowered the manure P content slightly (III). Therefore, as N was the limiting factor when applying dairy ma- nure on ley, separation had no influence on dairy manure application rates (Table 19). Sedimentation was a more efficient means to lower the swine manure solids and P content (I), contributing signifi- cantly to the permitted fertilizer appli- cation rates. Nitrogen being the limiting factor, slurry manure application rate was increased from 75 tn/ha to over 112 tn/ ha (Table 19). However, as stated in pre- vious studies (e.g. Ndegwa et al., 2001), slurry dry matter content has a strong in- f luence on the obtained sedimentation result, and as manure solid content is in- creased over 2% the sedimentation re- sult is degraded. In the present study, the used swine manure differed considerably in its average slurry manure TS compo- sition (3.5%, Table 2). Thus, in practice the requisite TS reduction might be dif- ficult to reach by sedimentation. Biological aeration treatment – odor emis- sions and hygienic quality of manure Biological aeration treatment influenced in particular manure odor (I) and hygienic quality (IV). Odor has been a major factor causing complaints and it has a particular relevance to environmental permit issues concerning barn expansions or construc- tion of new buildings. Odor emissions can be significantly reduced by reducing the odor potential of slurry manure. Prelimi- nary results of swine manure showed that biological treatment reduced the odor in- tensity to a level of no odor or only a very faint odor (I). The aeration treatment was probably able to reduce the odor emis- sions originating from the manure stor- age. The odor problem is particularly in- tense at the time of application and this approach would greatly facilitate odor re- duction. This study did not reveal wheth- er the achieved odor reduction continues after storing (aerated slurry). This issue re- quires further research. The other major advantage was that the slurry after aeration treatment had a clear- ly reduced content of pathogenic organ- isms (IV). Although manure storage has a hygiene improving influence, pathogens may survive for long periods. Pathogens are likely to survive viable even longer, for several months or even years, in soils where they are protected from exposure to UV radiation and desiccation (e.g. Nicholson et al., 2005). Biological treatment was observed to re- duce the viscosity, improving the pumping properties of slurry. This can probably also affect the spreading of manure on fields by improving the infiltration of effluent into the soil and providing for irrigation of ef- fluent. On the other hand, the high pH of aerated slurry manure may promote am- monia emissions during manure spread- ing. After biological treatment, nitrogen remained the limiting factor and applica- tion rates were only increased slightly to 130 t ha-1 at the best (Table 19). MTT SCIENCE 29 73 The major drawback related to aeration treatment is the required energy con- sumption (Westerman and Zhang, 1997). Therefore, it is essential to relate the aer- ation to the actual oxygen need based on real time ORP measures in treatment tanks (III). This enables the achievement of good treatment results by using low aer- ation rates, as was shown in the present study (I, III). By insulating the treatment reactors, the heat energy released during the process has been used for hygieniza- tion (Juteau, 2006). The captured heat en- ergy can alternatively be exploited by us- ing heat pump technology for e.g. heating the farm buildings. In the present study nitrous oxide (N2O) emissions were not measured. Normally nitrous oxide (N2O) is formed by biolog- ical denitrification under anaerobic con- ditions, but N2O as well as nitric oxide (NO) can also be formed as a by-product of the microbial nitrification. Béline et al. (1999) reported that relatively high N2O emissions were observed during aerobic biological treatment of livestock effluents (up to 20% of the total N). However, these emissions could be reduced to almost zero if good treatment conditions were applied (Béline and Martinez, 2002; Loyon et al., 2007). Based on the observed carbon re- ductions (Table 17), it was evident that carbon dioxide emissions occurred during aeration treatment. However, with limited aeration, the effort was to curb substantial carbon losses (I). Further research is need- ed to optimize the aeration level with re- gard to avoiding nitrogen oxide emissions while evading substantial carbon dioxide emissions and maximizing the heat ener- gy released. Air stripping – N removing The obtained results showed that, regard- ing manure application rates as fertilizer, the greatest benefits were achieved already during pre-treatment of solids separation, which clearly reduced the manure P con- tent (Table 19). Continuing the treatment with biological aeration and thereafter by sequential stripping allowed efficient ni- trogen separation, reducing the manure N content by 70% (II) and enabling the ma- nure N fixation into mineral form. The N- and P-poor reject can, for example, be sprinkled, on the fields nearby a farm using a high application rate (260 t/ha, Table 19). 74 MTT SCIENCE 29 Ta b le 1 9 N ut rie nt c on te nt s of s ep ar at ed f ra ct io ns (I − III , S up p le m en ta ry d at a) a nd c or re sp on d in g m an ur e ap p lic at io n ra te s w he n m an ur e is u til iz ed a s a fe rt ili ze r fo r si - la ge le y in t he e st ab lis hm en t y ea r. M an ur e Fr ac ti on  T S N to t N H 4+ P t ot A pp lic at io n ra te , N li m it a) (t /h a) A pp lic at io n ra te , P li m it a) (t /h a) C ri te ri on a pp lic at io n ra te sb ) (k g/ ha ) (k g/ to n m an ur e) D ai ry R aw m an ur e 44 .0 2. 2 0. 40 78 90 N 1 70 P 36 II I Se pa ra te d liq ui d 2. 2 0. 35 78 10 3 Se pa ra te d so lid Af te r a er at io n& str ip pi ng n o M gO 1. 0 0. 23 16 2 15 7 Af te r a er at io n& str ip pi ng +M gO 1. 2 0. 09 14 6 40 5 Sw in e R aw m an ur e 18 .5 2. 3 1 .6 0. 45 75 80 N 1 70 P 36 I Sl ur ry a fte r s ed im en ta tio n 5. 5 1. 5 1. 4 0. 08 11 3 44 4 Af te r a er at io n 4. 9 1. 5 1. 3 0. 05 11 4 70 6 II Sl ur ry a fte r s ed im en ta tio n 10 .1 1. 8 1. 5 0. 25 94 14 3 Af te r a er at io n 7. 0 1. 5 1. 3 0. 17 11 0 21 7 Sl ur ry a fte r s ed im en ta tio n* 9. 9 1. 9 1. 6 0. 49 87 73 Af te r s tr ip pi ng 7. 5 0. 7 0. 5 0. 04 26 0 85 0 Su pp l. Sl ur ry a fte r s ed im en ta tio n 6. 0 1. 5 1. 3 0. 10 11 0 37 4 da ta Af te r a er at io n 5. 1 1. 3 1. 2 0. 07 13 0 55 7 a) Ap pl ic at io n ra te a cc or di ng to th e N a nd P c on te nt s o f t he sl ur ry m an ur e at d iff er en t t re at m en t s ta ge s b) C rit er io n ap pl ic at io n ra te s o f N a nd P fo r l ey a cc or di ng to th e Ag ri- En vi ro nm en ta l p ro te ct io n Ac t. P ra te s o n le y on e sta bl ish m en t y ea r w he n so il P- va lu e is fa ir an d le y is so w n w ith c er ea l *c al cu la te d m ea n fro m th e se qu en ce st rip pi ng p ro ce du re re su lts II : T ab le 2 . MTT SCIENCE 29 75 6 Conclusions A new biological manure treatment scheme was proposed and evaluated on a pilot scale. The treatment consisted of a bi- ological treatment operated in a specially designed reactor regime using 600-L tanks followed by ammonia separation conduct- ed by ammonia stripping. Biological treatment in a series of contin- uously fed tanks applying limited aeration accelerated humification of the organic matter of slurry manure and increased the manure pH. Reduced odor and improved hygienic quality of slurry manure were also among the achieved results of the treat- ment. The designed system was operated effectively with HRT (hydraulic retention time) being 3 to 7 days when the DM (dry matter) content of manure was reduced by pretreatment to below 2%. The optimum redox profile of a stable system followed an increasing trend. It was shown experimentally that the de- signed treatment system can be loaded with a 3 to 4 days HRT value, if the DM content of the slurry manure is low enough (in swine manure less than 1%). The treat- ment system was loaded with mechanical- ly separated dairy slurry manure (DM 1.5 to 2%). In this case, the system could be loaded with a maximum of 350 l per day, corresponding to 6.8 days HRT-value. The results obtained suggest keeping the re- dox values of the treatment tanks ideal- ly at zero or slightly above in the 1st tank, at around 100–180 in the 3rd tank and at 180–200 mV in the 6th tank. It was shown that several benefits could be achieved when using a series of contin- uously fed aerated tank reactors and feed- back. The serial system proved to be sta- ble and resistant to process disturbances, and process failure was easily predictable and thus easy to control. The efficiency of the treatment increased by feedback via a number of tentative mechanisms: feedback dilution probably enhanced carbon reduc- tion and feedback probably stabilized and enhanced microbial function in the first tank. Due to limited aeration and feedback, some changes in the carbon and nutrient contents of the slurry manure were ob- served during biological treatment. Car- bon reduction varied in different stud- ies between 11% and 57% and depended on the initial carbon content. Nitrogen changes occurred mainly due to ammo- nia volatilization and varied from 0 to 15 % in swine manure. In dairy manure, an average of 52% maximum decrease was observed when ammonia-reduced feed- back was conducted after air stripping. Ni- trate formation was low with the aeration rates used. The concentrations of the to- tal phosphorus and divalent ions were also decreased. This was suggested to be due to the humic substances formed during bio- logical treatment. Humic substances be- have like weak acid polyelectrolytes and the occurrence of anionic charged sites ac- counts for the ability to retain cations. The most marked advantage of the biologi- cal treatment was its effect on odor and hy- giene. Odors were reduced efficiently to an insignificant level or only very faint odor was observed after four days of treatment. It was also shown that the six tanks in se- ries configuration with feedback served as a good means to achieve a good hygiene result of the treated manure. At best, over 76 MTT SCIENCE 29 90% reduction was observed in the num- bers of enteric indicator organisms. It was shown that the buffer system re- sisting pH change in slurry manure was composed of TAN (total ammoniacal ni- trogen), CO2, HCO3- and CO32−and can be circumvented by making use of the pH increase obtained by biological treatment, thus reducing the chemical consumption. Over 30% TAN removal by air stripping was shown to be possible without chemi- cal use, if the pH of the biologically treat- ed swine manure was above 8.9. It was shown that biological aeration treatment and thereafter N separation conducted by sequential stripping allowed efficient ni- trogen separation by reducing the manure N content by 70%. In the future, it would be interesting to study the preparation of seeding material, the effect of seeding and changes in mi- crobial community composition in a serial system in more detail using the latest an- alyzing tools available (e.g. using 454 py- rosequencing technology). It would also be interesting to study the changes be- tween the different treatment tanks on a compounds level (e.g. using metabolomic techniques) and focusing on humification to examine changes in molecular sizes and relationships with odor reduction. As a final conclusion of this study, the im- portance of deep understanding of the var- ious biological and physicochemical factors involved in complex raw material cannot be overestimated when developing manure treatment technologies. Combination of these factors may provide new insights in development work and provide advantag- es (such as reducing chemical consumption and energy demand) in manure treatment technologies. MTT SCIENCE 29 77 7 References Abu–Reesh, I. 2010. Optimal design of mul- ti–stage bioreactors for degradation of phenolic industrial wastewater: Theo- retical analysis. Journal of Biochemical Techology 3: 175–181. Alaerts, G. J., Rahman Mahbubar, M. D. & Kelderman, P. 1996. Performance anal- ysis of a full–scale duckweed–covered sewage lagoon. Water Research 30: 843–852. Angelidaki, I. & Ahring, B. K. 1993. Thermo- philic anaerobic digestion of livestock waste: the effect of ammonia. Applied Microbiology and Biotechnology 38: 560–564. Angenent, L. T., Sung, S. & Raskin, L. 2002. Methanogenic population dynamics during startup of a full–scale anaero- bic sequencing batch reactor treat- ing swine waste. Water Research 36: 4648–4654. Arnold, M., Kuusisto, S., Wellman, K., Kajol- inna, T., Räsänen, J., Sipilä, J., Puumala, M., Sorvala, S., Pietarila, H. & Puput- ti, K. 2006. Hajuhaitan vähentäminen maatalouden suurissa eläintuotan- toyksiköissä. VTT Tiedotteita 2323. Es- poo, Finland. 74 p. Arogo, J., Zhang, R. H., Riskowski, G. L., Christianson, L. L. & Day, D. L. 1999. Mass transfer coefficient of ammonia in liquid swine manure and aqueous solu- tions. Journal of Agricultural Engineering Research 73: 77–86. Arvidsson, J., Sjöberg, E. & Van den Akker, J. J. H. 2003. Subsoil compaction by heavy sugarbeet harvesters in southern Sweden III. Risk assessment using a soil water model. Soil & Tillage Research 73: 77–87. Ashley, K., Mavinic, D. & Koch, F. 2009. International Conference on Nutrient Recovery From Wastewater Streams Vancouver. London, UK: IWA Publish- ing. 904 p. Åström K. J. & Murray, R. M. 2009. Feed- back Systems: An Introduction for Sci- entists and Engineers. Princeton NJ, USA: Princeton University Press. 408 p. Aust M. O., Thiele– B. S., Eckhardt K. U. & Leinweber P. 2009. Composition of organic matter in particle size fraction- ated pig slurry. Bioresource Technology 100: 5736–5743. Avnimelech, Y. & Laher, M. 1977. Ammonia volatilization from soils. Equilibrium con- siderations. Soil Science Society Ameri- can Journal 41: 1080–1084. Banister, S. S. & Pretorius W. A. 1998. Op- timisation of primary sludge acidogenic fermentation for biological nutrient re- moval. Water SA 24: 35–41. Béline, F. & Martinez. J. 2002. Nitrogen transformations during biological aer- obic treatment of pig slurry: effect of intermittent aeration on nitrous oxide emissions. Bioresource Technology 83: 225–228. Béline, F., Martinez, J., Chadwick, D. R., Guiziou, F. & Coste, C. M. 1999. Factors affecting nitrogen transformations and related nitrous oxide emissions from aerobically treated piggery slurry. Jour- nal of Agricultural Engineering Research 73: 235–243. Bernal, M. P., Alburquerque, J. A. & Moral, R. 2009. Composting of animal ma- nures and chemical criteria for compost maturity assessment. A review. Biore- source Technology 100: 5444–5453. Bhaduri, S., Wesley, I. R. & Bush, E. J. 2005. Prevalence of pathogenic Yers- inia enterocolitica strains in pigs in the United States. Applied Environmental Microbiology 71: 7117–7121. Bicudo, J. R. & Goyal, S. M. 2003. Patho- gens and manure management sys- tems: a review. Environmental Technol- ogy 24: 115–130. Black, J. G. 2012. Microbiology Principles and Explorations 8th edition. Hoboken, NJ, USA: John Wiley & Sons Inc. 968 p. 78 MTT SCIENCE 29 Block, S. S. 2001. Disinfection, Sterilization, and Preservation. 5th Edition. Philadel- phia, PA, USA: Lippincott Williams & Wilkins. 1481 p. Bock, E., Koops, H. P. & Harms, H. 1989. Nitrifying bacteria. In: Schelgel, H. G., Bowien, B. (eds.). Autotrophic Bacte- ria. Berlin, Germany: Springer–Verlag. p. 81–96. Bonmati, A. & Flotats, X. 2003. Air strip- ping of ammonia from pig slurry: char- acterisation and feasibility as a pre– or post–treatment to mesophilic anaero- bic digestion. Waste Management 23: 261–272. Brown, T. E., LeMay, H. E. H., Bursten, B. E., Woodward, P., Murphy, C. 2011. Chemistry The Central Science 12th Edi- tion. Prentice Hall. 1195 p. Böhm, R. 2008. Survey on available meth- ods and techniques for the inactiva- tion of pathogens in animal manure. In: Koutev, V. et al. (eds.). Proceedings of the13th International Conference of the FAO ESCORENA Network on Recycling of Agricultural, Municipal and Industrial Residues in Agriculture. Ramiran 2008, 11 – 14 June, Albena, Bulgaria. p. 27–31. Bøtner, A. & Belsham, G. J. 2012. Virus survival in slurry: Analysis of the stabil- ity of foot–and–mouth disease, classical swine fever, bovine viral diarrhoea and swine influenza viruses. Veterinary Mi- crobiology 157: 41–49. Burns, R. T. & Moody, L. B. 2002. Phos- phorus Recovery from Animal Manures using Optimized Struvite Precipitation. Proceedings of Coagulants and Floc- culants: Global Market and Technical Opportunities for Water Treatment Chemicals, May 22 – 24, 2002, Chica- go, Illinois. p. 1–7. Burns, R. T, Moody, L. B., Celen, I. & Bu- chanan, J. R. 2003. Optimization of phosphorus precipitation from swine manure slurries to enhance recovery. Water Science Technology 48: 139–46. Burton, C. H. 2007. The potential contribu- tion of separation technologies to the management of livestock manure. Live- stock Science 112: 208–216. Carpenter, S. R., Caraco, N. F., Correll, D. L., Howarth, R. W., Sharpley, A. N. & Smith, V. H., 1998. Nonpoint pollution of surface waters with phosphorus and nitrogen. Ecological Applications 8: 559–568. Carter, S. D. & HaeJin, K. 2013. Technolo- gies to reduce environmental impact of animal wastes associated with feeding for maximum productivity. Animal Fron- tiers 3: 42–47. CEN 2003. Air quality: Determination of odour concentration by dynamic olfac- tometry. EN 13725. European Commit- tee for Standardization, Brussels, Bel- gium. 70 p. Chang, R. 2000. Physical chemistry for the chemical and biological sciences. Sau- salito, CA: University Science Books. 1018 p. Chen, Y., Cheng, J. J. & Creamer K. S. 2008.Inhibition of anaerobic digestion process: a review. Bioresource Tech- nology 99: 4044–4064. Chen, S., Liao, W., Liu, C., Wen, Z., Kin- caid, R. L., Harrison, J. H., Elliott, D. C., Brown, M. D., Solana, A. E. & Stevens, D. J. 2003. Value–Added Chemicals from Animal Manure. Final Technical Report. PNNL–14495. Richland, WA: Pacific Northwest National Laboratory. 135 p. Christensen, M. L., Hjorth, M. & Keiding, K. 2009. Characterization of pig slurry with reference to flocculation and separa- tion. Water Research 43: 773–783. Cliver, D.O. 2009. Disinfection of animal manures, food safety and policy. Bi- oresource Technology 100: 5392–5394. Conroy, P. J. 1997. Rensselaer Polytech- nic Institute (RPI). School of engineer- ing. Study material of the Department of chemical and biological engineering. Available at http://www.rpi.edu/dept/ chem–eng/Biotech–Environ/IMMOB/re- cycle.htm Accessed 31 July 2014. Craggs, R. J., Heubeck, S., Lundquist, T. J. & Benemann, J. R. 2011. Algae biofuel from wastewater treatment high rate algal ponds. Water science and tech- nology 63: 660–665. DaSilva, E. J. 1979. Biogas generation: de- velopments, problems and tasks; an overview. In: Bioconversion of organic residues for rural communities (Pa- pers presented at the Conference on the State of the Art of Bioconversion of Organic Residues for Rural Communi- MTT SCIENCE 29 79 ties, Nov 13 – 15 1978, Guatemala City, Guatemala). Food and Nutrition Bulletin, Supplement (UNU), no. 2. Paris, France: Division of Scientific Research and High- er Education, Unesco. p. 84–98. Dao, T. H., Sikora, L. J., Hamasaki, A. & Chaney, R. L. 2001. Manure phosphorus extractability as affected by aluminum– and iron by–products and aerobic com- posting. Journal of Environmental Qual- ity 30: 1693–1698. De–Bashan, L. E. & Bashan, Y. 2004. Re- cent advances in removing phosphorus from wastewater and its future use as fertilizer (1997–2003). Water Research 38: 4222–4246. Diaz, L.F., de Bertoldi, M., Bidlingmaier, W., 2011. Compost Science and Technol- ogy. Elsevier. 380 p. ECETOC, 1994. Ammonia emissions to air in Western Europe. Technical Report No. 62. Brussels: European Centre for Ecotoxicology and Toxicology of Chemi- cals. 196 p. Emerson, K., Russo, R. C., Lund, R. E. & Thurston, R. V. 1975. Aqueous ammo- nia equilibrium calculations: Effect of pH and temperature. Journal of the Fish- eries Research Board of Canada 32: 2379–2383. Erickson, L. E., Fan. L. T. 1968. Optimization of hydraulic regime of activated sludge systems. Journal of the Water Pollution Control Federation 40: 345–362. Esaki, H., Morioka, A., Ishihara, K., Kojima, A., Shiroki, S., Tamura, Y. & Takahashi, T. 2004. Antimicrobial susceptibility of Salmonella isolated from cattle, swine and poultry (2001–2002): report from the Japanese Veterinary Antimicrobial Resistance Monitoring Program. Jour- nal of Antimicrobial Chemotherapy 53: 266–270. European Commission, 2010. Eurostat. Pig farming statistics.