MultiBio: Environmental services from a multipurpose biorefinery

MultiBio: Environmental services from

a multipurpose biorefinery

Authors:

Sudhanshu S. Pawar, Alan Werker, Simon Bengtsson, Maria

Sandberg, Markus Langeland, Magnus Persson, Karin Willquist

Financed by

Vinnova

Content

Summary ... 3

Sammanfattning ... 4

Background ... 5

Project Results and Developments ... 7

Biohydrogen and acetate production ... 7

Production of polyhydroxyalkanoates (PHA) ... 9

Evaluation of Caldicellulosiruptor cells and PHB as feed components for Nile Tilapia fish ... 14

Feasibility study of the MultiBio concept ... 17

Communication and Dissemination ... 20

Summary

MultiBio project aimed to establish and demonstrate a novel multipurpose biorefinery cascade concept, producing three renewable biobased products: 1) biohydrogen, 2) biopolymers and 3) protein rich meal ingredients for fish farming. The cascade concept exploits the ability of a bacterium (Caldicellulosiruptor saccharolyticus) to transform nutrients present in low-value waste process waters of the pulp and paper industry, to high-value products hydrogen gas, organic acids and microbial biomass. The organic acid rich effluent will then be managed in an open culture microbial process used to achieve discharge water quality objectives and to produce polyhydroxyalkanoate (PHA) biopolymers. Moreover, since C. saccharolyticus protein content is more than 63% of cell dry weight, their potential in formulation of fish feed was evaluated.

A fiber sludge containing, CTMP residual stream was found to be a possible feedstock for the MultiBio process concept. Due to safety risks the demo-scale experiments of biohydrogen gas technology were moved from Biorefinery demo plant (Örnsköldsvik) of 40 m3 _{capacity to ATEX classified pilot-scale facility with 0.4 m}3 _{capacity. Hence, bacterial} biomass enough for the large-scale fish feed ingredient could not be produced. Lab-scale experiments with

Caldicellulosiruptor cells as fish feed ingredient showed promising results as a protein-rich, sustainable fish feed

ingredient. In addition, PHA biopolymer also showed favourable results as fish food ingredient for experiments at Gårdsfisk AB. Lab-scale experimental tests showed that the surplus activated sludge from the mills wastewater treatment could currently accumulate PHA to about 20 % of its dry weight. Mass balance evaluations based on realistically achievable expectations indicated a PHA biopolymer production potential of 3 600 tons of PHA per year from available organic residuals and for the two evaluated mills combined.

The MultiBio concept has a positive climate impact in comparison with current treatment and moves developments in a positive direction to achieve 7 of the 10 Swedish environmental goals. Through a detailed feasibility analysis, a natural progression in next steps in scenarios were suggested for PHA production. The MultiBio cascade process can be implemented with further necessary development with good business potential and a positive effect on climate change. However, biohydrogen technology needs further developments before this cascade process concept can be implemented. Alternatively, a scenario with only biopolymer technology shows already a significant business potential and even larger positive effect on climate change. A successful next step in demonstration of the PHA biopolymer production scenario may lead to it being implemented within the next few years. Furthermore, MultiBio has attracted a lot of attention regionally and nationally but also internationally with a total of 65 media listings. A licentiate thesis and three university degree projects linked to the project have been completed. Overall, the MultiBio project has successfully achieved its goals and objectives.

Sammanfattning

MultiBio syftade till att etablera och demonstrera ett nytt bioraffinaderi-kaskadkoncept med tre förnybara biobaserade produkter: 1) bioväte, 2) biopolymerer och 3) proteinrika foderingredienser för fiskodling. Kaskadkonceptet utnyttjar förmågan hos en bakterie (Caldicellulosiruptor saccharolyticus) att omvandla näringsämnen som finns i massa- och pappersindustrins lågvärdiga processavloppsvatten till högvärdiga produkter vätgas, organiska syror och mikrobiell biomassa. Det utgående vattnet, rikt på organiska syror, hanteras sedan i en bioprocess med blandad mikrobiell kultur som används för att rena processvattnet och samtidigt producera biopolymerer av typen polyhydroxyalkanoater (PHA). Eftersom C. saccharolyticus proteininnehållet är mer än 63 % av celltorrvikt, utvärderades deras potential för beredning av fiskfoder.

En fiberslam-innehållande CTMP-restström visade sig vara en lämplig råvara för konceptet. På grund av säkerhetsrisker flyttades demoskalaexperimenten av biovätgasteknik från Biorefinery-demoanläggning (Örnsköldsvik) med 40 m3 kapacitet till ATEX-klassificerad pilotskaleanläggning med 0,4 m3 _{kapacitet. Därför kunde inte tillräckligt med} bakteriebiomassa för den storskaliga fiskfoderingrediensen produceras. Experiment i laboratorieskala med

Caldicellulosiruptor-celler som fiskfoderingrediens visade lovande resultat som en proteinrik, hållbar

fiskfoderingrediens. Dessutom visade PHA-biopolymeren gynnsamma resultat som fiskfoderingrediens för experiment på Gårdsfisk AB. Experimentella test i laboratorieskala visade att bioslammet från bruken kunde ackumulera PHA till cirka 20 % av dess torrvikt. Massbalansbedömningar baserade på realistiska förväntningar indikerade en produktionspotential på 3 600 ton PHA per år från tillgängligt organiskt avfall vid de två ingående bruken.

MultiBio-konceptet har en positiv klimatpåverkan jämfört med nuvarande behandling och har potential att bidra i rätt riktning för att uppnå 7 av de 10 svenska miljömålen. Genom en detaljerad genomförbarhetsanalys föreslogs scenarier med en stegvis implementering. MultiBio-kaskadprocessen kan implementeras med ytterligare nödvändig utveckling med god affärspotential och en positiv effekt på klimatförändringen. Men bioväte-tekniken behöver vidareutvecklas innan detta kaskad-koncept kan implementeras. Samtidigt visar ett scenario med enbart biopolymerteknologi redan nu en signifikant affärspotential och ännu större positiv effekt på klimatförändringen. En framgångsrik demonstration av det senare scenariot med endast PHA-produktion kan leda till att det genomförs inom de närmaste åren. Dessutom MultiBio har rönt stor uppmärksamhet regionalt och nationellt men även internationellt med totalt 65st medianoteringar. En licentiatavhandling och tre examensarbeten har färdigställts kopplat till projektet. Sammantaget har MultiBio framgångsrikt uppnått sina syften och mål.

Background

MultiBio aims to evaluate a multipurpose cascading biorefinery with the capacity to provide environmental benefits from producing three products in a cascade process: 1) biohydrogen, 2) biopolymers, and 3) sustainable feed ingredients for farmed fish (Fig. 1). The concept exploits the ability of a bacterium (Caldicellulosiruptor

saccharolyticus) to transform hardwood and/or softwood derived hemicellulose and nutrients present in the effluent

process water from the pulp and paper industry to renewable products (1 and 3) and to exploit the intermediate

effluent in a subsequent bioprocess to produce biopolymers (2). The development of this cascade concept is motivated due to expected contributions to meet several of the Swedish environmental objectives - Reduced Climate Impact, Clean Air, Natural Acidification Only, A Non-Toxic Environment, Flourishing Lakes and Streams, A Balanced Marine Environment, Flourishing Coastal Areas and Archipelagos, and A Rich Diversity of Plant and Animal Life.

Figure 1 A schematic of the MultiBio concept.

Hydrogen gas (H2) is generally sourced today from natural gas, and climate impact of its use is expected to be improved by the introduction of renewable and sustainable hydrogen sources. An environmentally, economically, and socially sustainable method for biohydrogen production from pulp and paper industry waste streams was a primary focus as part of the project plan. The work conducted on the hydrogen producing bacterium C. saccharolyticus has gained attention in recent years. In part, this attention reflects the hydrogen production potential of the microorganism where the theoretical hydrogen yield of 4 mol/mol hexose sugars has been demonstrated. Further, the bacterium displays a preference for pentose sugars, abundant in the hemicellulose fraction of e.g. in pulp and paper industry residue effluent streams. Moreover, as C. saccharolyticus needs thermophilic conditions of 70 °C, surplus mill heat was considered as means to warm the reactors and/or the residual stream in a full-scale plant.

The next technology in the cascade process is biopolymer production. It is based on the utilization of acetate rich effluent production by C. saccharolyticus. Acetate, and volatile fatty acids in general, are readily metabolized by many bacteria and converted into intracellular biopolymers. The state-of-the-art is with open mixed culture methods and processes for biopolymer production concomitant to the process effluent water quality management. These biopolymers are from the family of polyhydroxyalkanoates (PHAs), and they can be recovered and formulated into completely biodegradable thermoplastics with potential applications also within the pulp and paper industry such as for example coatings and wood-fibre biopolymer composites. This component of the cascade process was also envisioned to contribute to

Current low-value streams to Tomorrow's Bio-products

Biological treatment

Bio-H

Bio-H2– Biohydrogen process at 70 C PHA – Polyhydroxyalkanoates process

VFA – Volatile fatty acids Microbial biomass

earmarked national environmental objectives. Industrial effluent water quality management methods have been shown to be integrated with PHA-rich biomass production with up to 70 % gPHA per gram dried biomass under real operating environments at pilot scale. The recovered PHAs have also been demonstrated to meet commercial standards as ingredients for bioplastic formulations. The ‘zero waste’ philosophy of these developments provides for biological treatment of process waters while recovering renewable resources as biopolymers, lipids, minerals, other platform chemicals and energy.

The pivotal objective of a cascade biorefinery is the goal of feeding a ‘sustainable circular economy’ such that a low-value waste stream of one process is transformed into renewable resources as feedstocks for downstream processes. Previous studies have shown that the C. saccharolyticus cells are rich in protein (~63% proteins/cell dry weight). Moreover, the amino acid profile corresponds well with the amino acid requirements of Salmonid species. Thus, microbial biomass from biohydrogen process has anticipated potential as a feed ingredient in aquaculture diets and was trialled as fish-feed ingredient.

The cascading use of residuals from different industrial processes and the use of low value spill heat offer to create an industrial symbiosis with pulp and paper industry activities. This would not only enable for circularity of products and by-products but also advance in developments that contribute to wider environmental objectives.

The project was financed by Vinnova (Dnr – 2017-03286) and co-ordinated by RISE Research Institutes of Sweden AB. MultiBio stimulated wider collaboration and interests. To begin with MultiBio embraced scientists and experts from Lund university, Karlstad University, Swedish agricultural University (Uppsala), Paper province, and RISE. It also engaged with three pulp and paper industries – Stora Enso Skoghall AB, Rottneros Bruk AB and BillerudKorsnäs Sweden AB, who were potential stakeholders to bring the concepts under investigation to future practice. Together with the mills, Drinor AB was also seen as a stakeholder to implement the technology in combination with their innovative pressing methods. Experts from Fortum Recycling and Waste AB, Promiko AB contributed with their expertise in development and implementation of the concept elements. Gårdsfisk AB provided links to application as local and market leaders in pisciculture.

Project Results and Developments

Biohydrogen and acetate production

Five different strains of Caldicellulosiruptor originally purchased from DSMZ, Germany were kindly obtained from the Division of Applied Microbiology, Lund University and were used in the project (Table 1).

Table 1 Strains of Caldicellulosiruptor used in the experiments

Strain DSMZ # Taxonomical name

DSM-8903 Caldicellulosiruptor saccharolyticus

DSM-6725 Caldicellulosiruptor bescii

DSM-18902 Caldicellulosiruptor kronotskyensis

DSM-13100 Caldicellulosiruptor owensensis

DSM-18901 Caldicellulosiruptor hydrothermalis

First we performed experiments at lab-scale where we assessed the tolerance to different residual streams, namely – ‘fibre-rich, CTMP residual stream’, TMP stream, ‘blekeri’ bleaching process residual stream from pulp and paper mills and a residual stream obtained from a pressing process. Experiments were performed in flasks as well as in controlled CSTR bioreactors. The experiments in CSTRs were performed by feeding the influent in semi-continuous mode to a co-culture of five chosen Caldicellulosiruptor species (Table 1). The results showed that a residual fiber concentrated sludge containing mill CTMP process effluent did not cause inhibition to growth or H2 production. Thus, this particular process effluent was deemed to be a suitable candidate for further experiments towards developing methods for a biohydrogen production process at a pulp and paper mill.

Figure 2 The bioreactrors used for lab-scale experiments for bihydrogen production using a fibresludge containing, CTMP effluent stream

In further experiments an Upflow-anaerobic (UA) reactor (Fig. 2) containing K1-carriers and the fibre sludge also as carrier materials for biofilm formation, was fed with undiluted fiber sludge containing CTMP effluent.

Caldicellulosiruptor species were able to ferment the sugars naturally present in the influent but experiments were not

able to demonstrate signs for hydrolysis and degradability of influent cellulosic fibres. The influent stream was found to be too dilute in soluble sugar concentration for biohydrogen production, and therefore concentrating this process

CSTR

UA

water was considered to be necessary in order to be able to increase production rates of a biohydrogen process. To test such a concentrated influent stream scenario, a synthetic medium containing hemi-cellulose derived sugars at a high concentration, was fed to the UA reactor which produced a maximum of 309 mL of H2 /L /h. Thus, a process treating more concentrated fibre sludge process waters are anticipated to be able to produce H2 at significant rates.

Up-concentration of the influent requires further evaluation via technoeconomic analysis towards understanding costs and benefits. An ability for the maintained hydrogen production at an HRT of 3.3 h provided promise to apply the

technology with smaller footprint. This outcome can be compared to biogas plants that can have hydraulic residence times between 12-24 h in fast operating systems.

Methods for handling, storage and growth of the obligate anaerobic bacteria were developed and benchmarked for larger scale applications. Urea was evaluated as an alternative N-source to meet the biohydrogen process added nutrient requirements. C. saccharolyticus is postulated to be the only strain able to utilize urea as a nitrogen source, and urea for N supply will serve to maintain a selective pressure for C. saccharolyticus to be a dominating strain in the co-culture mixture due to exploited metabolic competitive advantages.

Figure 3 Process set-up at IndiEnz AB. The facilities were distributed in 3 different rooms demarked with dashed lines in black. The pictures of the harvest tank, filtration unit and centrifuge are merely representative.

Demo-scale trials were initially intended to be performed at the Biorefinery Demo Plant (BDP, Örnsköldsvik, Sweden). However, after detailed risk assessments, due to identified safety concerns for handling H2 gas, it was determined that these trials unfortunately could not be performed at BDP as was originally planned and anticipated. An alternative pilot-scale facility was utilized at IndiEnz AB. These facilities were previously used for pilot-scale experiments of biogas production and hence were ultimately found to be more suitable for the biohydrogen demonstration trials. However, essential modifications were made before the trials were undertaken. More than 20 batch reactor experiments were performed in sequence in a 440 L reactor at 70 °C.

Scale-up from lab-scale to pilot-scale did not pose challenges for successful cultivation of Caldicellulosiruptor. To the best of our knowledge, this effort represented the largest scale at which Caldicellulosiruptor species have been grown. The methods used are scalable to even larger process volumes that become industrially relevant. More than 2 kg of dry cells of Caldicellulosiruptor species were delivered to SLU to perform experiments with fish feed formulations (Fig. 4).

The partial but significant success of the biohydrogen process was important for the project to reach its milestones and its potential toward the project environmental goals as were introduced. With this report highlight outcomes that have identified at least one principal process effluent that was found to be promising for business case evaluations for a biohydrogen production process. Much has been learned with respect to needs for strategy and considerations in handling risks that come with hydrogen production. Initial steps towards increasing the productivity with higher

440 L 60L Caldi cells to WP4 Harvest tank Filtration Centrifugation LU ATEX-class room • Reactors • Medium tank • N2supply tank Control room • pH controllers • Water bath at 78 °C • Pumps • Harvest tank Process room Dry at 30-40°C

process flow rate through the application of biofilm principles were made. These principles are considered to be central towards the continued development and implementation of an effective and viable scaled up biohydrogen production process.

Figure 4 Last batch of dried cells of Caldicellulosiruptor species shipped to SLU

Overall, through this work, the biohydrogen process using Caldicellulosiruptor species was scaled-up to TRL-5. Although results were promising in nature, clearly, the biohydrogen process remains a few steps away from conceptualising implementation in a full-scale industrial production plant.

Production of polyhydroxyalkanoates (PHA)

As part of the MultiBio project, the potential for production of biobased and biodegradable polymers, from organic residuals in process effluent streams, was evaluated for Stora Enso Skoghalls and Rottneros pulp and paper mills in Värmland, Sweden. These biopolymers are thermoplastic polyesters made by naturally occurring bacteria and are from the family of polyhydroxyalkanoates (PHAs). They can be formulated as principal ingredients for plastics, functional chemicals, and composite materials of interest to a diversity of sectors including the pulp and paper industry. They can also be formulated into fish feed to impart pre/pro-biotic benefits in aquaculture. Commercial quantities and qualities of PHAs can be produced as a corollary benefit to biological wastewater treatment processes that are used to treat industrial and municipal wastewaters for environmental protection. The goal of this part of the present study has been to determine the potential quantities of PHA that could be produced as an integral part of the mill residual organics and effluent management. The aim of this MultiBio work package has also been to recommend steps forward. Details of methods and findings presented in this final report are available in a comprehensive report that was generated specifically for this work package1

It was estimated that Skoghalls and Rottneros mills manage 70 and 15 tons per day of residual organic material on a chemical oxygen demand (COD) basis, and in relation to respective production levels of about 778,000 and 170,000 ton/yr paper and board. These numbers were calculated from mass balances (Figures 1 and 2) using historical monitoring information provided by the mills and with measurements made in this study on samples from selected locations of the treatment processes. The residual COD is comprised of fibre and surplus activated sludge as well as soluble dissolved organic material. This organic material is a resource to produce a microbial biomass rich in PHA with

1_{Werker, A. and Bengtsson, S. (2020). The production of biopolymers for bioplastics using pulp and paper mill wastewater and}

remaining organic material being used for boiler heat production. Improved sludge dewatering is expected such that COD used for heat production today could be diverted to PHA without undue loss of heat production capacity in the balance.

Qualitative assessments of the surplus activated sludge from the mills indicated for some degree of selection for the PHA storing phenotype in the biomass (Figure 3). However, in subsequent quantitative experimental evaluations of the respective mill surplus activated sludge (Figure 4), it was found that neither Skoghalls nor Rottneros currently generate activated sludge with sufficient PHA accumulation potential today. The biomass could accumulate approximately 20 % of final organic weight as PHA. Notwithstanding that methods to improve the selection and performance of the biomass for PHA storage do not need to require onerous modifications and upgrades of process, the most prudent way to progress for PHA production would be, to start, to use the waste activated sludge as an organic substrate rather than as an active biomass (Werker and Bengtsson, 2020).

Therefore, two process scenarios were proposed as a logical progression and a strategic way forward (Scenario 1 & 2, Figure 5). PHA production requires that the organic feedstock is first converted into volatile fatty acids (VFAs) through acidogenic fermentation. If all the residual organic mass is used to produce VFAs (Scenario 1), it was estimated that about 2 000 and 400 tons PHA per year could be produced at Skoghalls and Rottneros mills, respectively. Production amounts would be increased (Scenario 2) by about 50 % (3 000 and 600 tPHA/yr, respectively) if the residual mass flows of waste activated sludge biomass from the mill aerobic biological wastewater treatment processes could be exploited for a PHA storing potential rather than just being an organic feedstock to yield VFAs.

The Scenario 2 mass balance evaluations were based on realistic expectation that such a waste activated sludge could be made to accumulate PHA up to 60 % of final organic weight. Thus, the performance of the waste activated sludge for PHA storage will then need to be improved from the present-day level of 20 % through careful attention to details of the bioprocess that influence selective pressures for enrichment of the PHA storing microorganisms in the activated sludge biomass. Such details can be subtle and without influence on the performance of the process in wastewater treatment and meeting effluent water quality objectives.

Even if it is in the realm of feasibility to manipulate selective pressures for improved PHA potential for the waste activated sludge today, such efforts were not considered to be strategic as a first step. A logical initial low risk development would be to first produce PHA using all the considered organic residuals just as an organic resource for VFAs (Scenario 1, Figure 5). Once an economically viable production activity for 2000 and 400 tPHA/yr was established, then the motivation to modify the existing wastewater treatment plants, to produce a better performing PHA storing biomass, would be well-motivated. An ability to further exploit the surplus activated sludge for its potential capacity to accumulate PHA rather than as just an organic feedstock for VFAs, would increase by 50 % the PHA production supply from the same mass flow of input material.

A polymer production activity for 2 000 and 400 tPHA/yr is small compared to the core revenue generating outputs from Skoghalls and Rottneros mills, being approximately 778 and 170 kton/yr pulp and board, or pulp, respectively. At the same time, assuming a niche polymer market exploitation at 5 €/kgPHA, a product revenue of 12 M€/yr may well support a worthwhile specialty industry. That supply chain for a worthwhile specialty industry is the lower risk opportunity to develop first. In the first steps, the symbiosis of PHA production feeding on mill organic residuals management should function with negligible if any burden on the mill daily routine in operations. At the same time, it should simplify or create otherwise harder to achieve benefits or further efficiencies to the current routine of mill operations that have been interpreted to be possible (Werker and Bengtsson, 2020).

A mill side-line of polymer production anchors in-house supply and experience with PHAs while casting a net for evolving and capturing future interesting materials and business. Future mill economies may foreseeably require standing on wider business footings including fibre as well as from other cellulose derived biobased revenue generating products and services. PHAs can naturally also be made using the prime input lumber starting with hemicellulose and

cellulose. Opening a PHA specialty industry first is a way to open the door for discovery that may well motivate directing strategically separated mill upstream organic fractions to successful large commercial revenues from PHAs in the future.

Figure 5 Outcome of COD balances for Skoghalls bruk for selected process effluent steams (2 years of data). Respective sampling

stations are indicated as circled nodes. Mass flows for soluble COD (sCOD), fibre COD (fCOD), and biochemical oxygen demand (BOD) are the estimated mean daily averages in tons oxygen per day.

Figure 6 Outcome of soluble COD and TSS assessments for COD estimations from Rottneros bruk based on provided process effluent steams (1 year of data). Respective sampling locations are indicated as circled nodes (A, B, C and D). Mass flows for

soluble COD (sCOD), fibre COD (fCOD), and active (biomass) COD (aCOD) are the estimated mean daily averages in tons COD per day.

Figure 7 Phase contrast and respective Nile Blue stained images of activated sludge after nominally 3-hour accumulation trials during week 38, 2018. Samples were from the Skoghall (top) and Rottneros (bottom) activated sludge processes. Red staining is

qualitatively suggestive of PHA accumulation by the microorganisms in the activated sludge as stimulated by the conditions of the testing. Magnification = 200 ×.

Figure 8 Parallel 450 mL bed-batch feed-on-demand activated sludge PHA accumulation reactors operated at room temperature with aeration providing mixing and oxygen supply. Dissolved oxygen was measured in the main well-mixed volume, and pH was

monitored in the recirculation flow. Volumetric cylinders provided substrate reservoirs. Air flow was regulated at constant supply pressure by rotometers with needle valves. Substrate in put was via diaphragm pumps supplying the green tubing shown in the figure.

Figure 9 Scenario 1 and 2 show summary flow diagrams depicting two feasible scenarios recommended as steps in progression to produce PHA from Skoghalls and Rottneros pulp and paper mill residual organic streams and to produce collectively from 2400 (Scenario 1) to 3600 (Scenario 2) tons of PHA/yr.

Evaluation of Caldicellulosiruptor cells and PHB as feed

components for Nile Tilapia fish

Within the project, two different feed experiments on Nile Tilapia (Oreochromis niloticus) have been performed. The first trial aimed to evaluate the protein rich microbial biomass from biohydrogen production and was performed at SLU in Uppsala. The aim of the second feed trial was to study the effects of dietary inclusion of the biopolymer PHB, on growth, immunology, gut health, and gut microbiome in Tilapia.

Dried Caldicellulosiruptor saccharolyticus was ground to a fine powder and mixed in a feed at three different inclusion levels (1, 3 and 6 % on a dry matter basis, Table 1) while replacing soy protein concentrate.

Table 2 Formulation of test diets (g/kg) for the first trial

Ingredients Control diet CP1 CP3 CP6

Fishmeal 10 10 10 10

Soy protein concentrate 10 10 10 10

Wheat gluten Wheat meal 10 18 10 18 10 18 10 18 Pot starch 13.5 13.5 13.5 13.5

Ingredients Control diet CP1 CP3 CP6 Fish oil

Rapeseed oil

Vitamin mineral premix Soybean meal Geletin Caldicells L-methionine

Titanium dioxide (TiO2)

Monocalcium phosphate CMC Total 3 5 1.5 18 8 0 - 0.5 1 1.5 100 3 5 1.5 16.5 8 1.5 - 0.5 1 1.5 100 3 5 1.5 15 8 3 - 0.5 1 1.5 100 3 5 1.5 12 8 6 - 0.5 1 1.5 100

The feeds were produced by pelletizing with warm conditioning. Tilapia "Rödstrimma" from Gårdsfisk AB, (97.2 ± 0.85 g; mean weight ± SEM) were divided into groups of 15 fish per tank (200 L; Figure 1). Each feed (treatment) was fed to three different tanks (triplicate), until the saturation for a total of 72 days.

At the end of the experiment, the fish had an average weight of 264.0 ± 3.89 g. A total of 4 fish died during the

experiment, but there was no connection found between treatment and mortality. Weight gain per treatment varied between 169-176 % (Table 2), and showed no statistical difference between diets. Similarly, the specific growth rate (SGR,% / day) varied between 1.27-1.30 % (increase / day, (table 2)) and did not show any statistical differences between diets.

Table 3 Growth performance, nutrient retention and relative body indices after 72 days

WG-weight gain (% per day), SGR- specific growth rate (% per day), FCR- feed conversion ratio, CP retention- crude protein % retained in the body, CL retention- crude lipid % retained in the body, HIS-hepato-somatic index, VSI-viscero-somatic index.

Furthermore, feed conversion ratio (FCR), crude protein (CP) retention, crude lipid (CL) retention, and viscero-somatic (VSI) index did not differ statistically between diets. However, fish fed diet with 5 % caldicell inclusion had statistically significant enlarged livers when compared to fish fed control and diet with 1 % inclusion level. The reason behind this liver effect is unclear but may indicate possible metabolic issues in these fish.

Diet WG (%) sem SGR

(%/dag) sem FCR sem CP retention % sem CL retention %

sem HSI sem VSI sem

CP1 170.5 5.07 1.28 0.02 1.0 0.05 53.24 2.84 147.63 17.98 1.32 0.07 8.37 0.35

PC3 168.8 2.84 1.27 0.01 1.1 0.05 47.75 1.60 148.24 26.02 1.62 0.08 7.63 0.48

PC5 170.7 5.03 1.28 0.02 1.2 0.08 46.58 2.54 145.35 5.20 1.87* 0.11 8.31 0.50

Ctrl 176.3 7 1.3 0.03 1.0 0.01 57.50 3.74 147.00 7.50 1.30 0.11 7.53 0.20

In the application step, our goal was to evaluate the digestibility of Caldicellulosiruptor cells in tilapia in the first experiment. Such experiments would include production of a reference diet without the test ingredient and a diet consisting of 30 % test ingredient and 70 % of the reference diet mixture. In such design, the diets are not balanced in terms of energy, protein and fat content, thus comparing growth performance is not applicable. However, such experimental design may provide insights in digestibility, taste (feed acceptance) and possibly other parameters such as intestinal health and effect on intestinal flora. In the experiment that was carried out however, it was not possible to assess digestibility due to low amounts of raw material (Caldicellulosiruptor cells) available.

Normally, new protein sources in fish feed are evaluated by including 10 to 50 %, and sometimes more, of the raw material in the feed which is often based on results from digestibility studies. The feed industry is normally more restrictive in the inclusion of new raw materials, and often includes from 5 % initially up to 20 % pending that the consumers do not respond negatively to the new raw material.

Based on the results of this trial, it can be concluded that soy concentrate can be replaced with Caldicellulosiruptor

saccharolyticus at levels 1-6 %, without any change in growth performance or feed intake of tilapia. This study would

need to be supplemented with a digestibility study, as well as trials with higher inclusion levels of Caldicellulosiruptor cells to establish upper inclusion limits. We did not examine the effect of Caldicellulosiruptor cells on intestinal health, which was also not planned in this project, but is needed for future applications.

The second trial was performed at Gårdsfisk AB in Åhus. In total 200 Tilapia “Svartstrimma” were fed diets

containing 0.0; 2.5; 5.0 and 7.5 % polyhydroxybutyrate (PHB; Table 3). In addition, a commercial diet was used as a positive control. The PHB (>98 %)) was supplied from a commercial producer (Biomer DE). This specific type of PHA is chemically the same as would be produced from acetic acid in the MultiBio cascading process concept. Feed was extruded at SLU.

Juvenile tilapia (59.6 ± 1.7 g; mean weight ± SEM) were randomly allocated to 20 experimental tanks (400 L) connected to a recirculating aquaculture system (RAS; Figure 2), in groups of 10 fish/tank. After 12 weeks of feeding fish were slaughtered, weighed and dissected. Samples were collected for analysis of growth, somatic indices (i.e. nutritional status), immunology, gut histology and gut microbiota.

Table 4 Formulation of test diets (g/kg) for the second feed trial

Ingredients Control diet PHB2.5 PHB5 PHB7.5

Fishmeal 5.0 5.0 5.0 5.0

Soy protein concentrate 4.0 4.0 4.0 4.0

Wheat gluten Wheat meal 14.5 18.5 14.5 16.0 14.5 13.5 14.5 10.2 Pot starch 10.0 10.0 10.0 10.0 Algae oil Rapeseed oil

Vitamin mineral premix Soybean meal Fava bean meal PHB

DL-methionine Titanium dioxide (TiO2)

Monocalcium phosphate Betain Total 5.0 1.5 1.7 20.0 16.0 0 0.5 0.5 1.8 1.0 100 5.0 1.5 1.7 20.0 16.0 2.5 0.5 0.5 1.8 1.0 100 5.0 1.5 1.7 20.0 16.0 5.0 0.5 0.5 1.8 1.0 100 5.0 1.5 1.7 20.0 16.0 7.5 0.5 0.5 1.8 1.0 100

Sds ds ds d

The final weight of the fish was 264.6 ± 5.3 g (mean weight ± SEM), with no statistical differences between groups. There was a numerical difference in the weight gain of the fish fed the commercial diet (319 %) and the experimental diets (average weight gain of 351 %), but not between any of the experimental diets. PHB does not have a positive effect on growth in juvenile Nile Tilapia. Further analyses will tell if positive effects of PHB is observed on the immune system, gut health or microbiota.

Feasibility study of the MultiBio concept

The MultiBio feasibility study was based on discussions with stakeholders, experimental results of biohydrogen production in lab and pilot scale, experimental studies of fish growth trials, calculated results of possible PHA production, an economical survey of product prices, and an LCA.

The production potential of biohydrogen, fish feed, fish and PHA-polymers has been estimated. Biohydrogen production is a central process in the MultiBio concept. The experimental studies showed that the concentration of sugar in form of cellulose and hemicellulose, and Caldicellulosiruptor -bacteria sensitiveness to some process effluents, limited efficient biohydrogen production to use fiber sludge.

A case study to estimate the mass flow of substrate and products at Skoghall Mill was used as a base for the feasibility study. Reject from the fiber sedimentation was used as substrate for the Caldicellulosiruptor -process. In the

Caldicellulosiruptor-process is biohydrogen, protein rich biomass for fish feed and acetic acid produced. The potential

of fish production was based on the mass of Caldicellulosiruptor-cells and the production of PHA-polymers was based on the mass of effluent acetic acid.

Since the Caldicellulosiruptor-process could use only fibre sludge containing CTMP effluent, it was limited to a small part of Skoghall´s total effluent; a second case was evaluated. In the feasibility study it is referred to as case Max PHA. With this technique, fiber sludge and bio-sludge are fermented under mesophilic conditions to be transformed into VFA. One third of the VFA is then used to build new bacteria with better PHA-accumulation properties than bio-sludge from the wastewater treatment plant. The rest of the VFA is then polymerized by the bacteria into PHA. With this concept almost all organic material in the Mill wastewater can be used to produce PHA. The disadvantage is that neither biohydrogen nor fish feed can be produced. This Max PHA concept has not been tested experimentally in this project. However, there are a lot of reliable references from scientific and practical experiences from which to build a model.

The economic potential was briefly evaluated by analyses of the price the different products can be sold for. The environmental impact of the MultiBio concept was evaluated with an E-LCA and discussion about environmental impacts was not possible to measure with an E-LCA. The results were then discussed according to seven of Sweden’s national environmental goals.

The overall feasibility of the MultiBio concept was then evaluated together with input from stakeholders.

Stakeholders input

There is a great interest among all involved stakeholders for the new bio-based products from the MultiBio concept. A pulp and paper mill is often a fairly closed system where most waste materials are used and recovered. Heat and some of the electricity needed for pulp and paper production are produced at the mills from waste materials. At Rottneros mill, the wastewater treatment plant is producing biogas to replace fossil energy needed in the pulp production. However, in the system, there are some waste materials with less value. Biosludge is difficult to dewater and it therefore has low energy value. It is of interest for all mills if bio-sludge can be used for better purposes. There are also large amounts of heat (40 - 60°C) that are wasted today.

Even if a pulp and paper mill is using mostly renewable materials, they still use some fossil energy. Today, Skoghall mill and Gruvön mill, use oil in the lime kiln. The oil could be replaced with biohydrogen gas. It is also possible to use biohydrogen as fuel for larger vehicles on the mill site.

Many paper packages are combined with a layer of plastic. There is also a big interest to see if PHA may replace polyethylene in carboard or for screw caps or straws needed for different types of packages. PHA polymers are biodegradable, also in marine environments. Properties of PHA are interesting for society and motivate developments for polymer manufacturers and application developers.

The unique protein combination in Caldicellulosiruptor cells makes them useful as a fish feed ingredient. The possibility to avoid soya protein or fish meal in fish feed, gives the feed less environmental impact and are therefore attractive to fish farmers like Gårdsfisk. However, fish feed manufacturers as Raisio, need secured supply chains of almost 1000 tons / year before they can consider using such new ingredients.

Mass flow analysis

In Figure 12, the results from the mass flow analysis are presented. With the MultiBio concept when fiber sludge reject is used as substrate for the Caldicellulosiruptor-process it possible to produce 37 tons of biohydrogen gas per year. Approximately 100 times this amount is needed to replace oil used in the lime kiln. 55 tons of protein rich biomass is also a much smaller amount than the fish feed producers would need commercially. With the MultiBio concept it is possible to produce 153 tons PHA-polymer per year. This was also considered as a relatively small amount for commercial purposes. However, with the Max PHA case, it is possible to produce more than 2000 tons of PHA polymer per year.

Economic potential

The estimated prices are listed in table 3.4.2 together with the possible income per year for each product. The price is presented with a span and the value used to calculate possible income is written in brackets.

Hydrogen gas can command a high price. However, the much larger production capacity for PHA ultimately give a higher income per year. For the Max PHA case was the potential income estimated to be more than 83 M SEK.

Table 5 Prices and possible income for the evaluated MultiBio concept.

Product Price SEK/kg Income MSEK / year

Hydrogen gas 85 - 125 (100) 4

Fish feed 10 – 15 (12) 1

Fish file´ 75 (50) 3

PHA polymer 30 – 50 (40) 6

E-LCA

The reference case corresponds to the biological wastewater treatment plant Skoghall mill was using in 2019. The environmental impact of petrol for Mill vehicles and fossil polyethylene was avoided with the new products hydrogen gas and PHA polymer. The MultiBio and Max PHA cases also use electricity and nutrients as in the reference case. However, avoiding fossil-based petrol and polyethylene have much larger positive environmental effects and climate change effect (Figure 13).

Figure 13 Environmental impact measured as a E-LCA for a reference case, only treatment plant, compared to the MultiBio and Max PHA concept.

Both Caldicellulosiruptor-process and PHA-polymer production require heat. In this E-LCA, soft wood chips were used as energy source for the heat. The use of soft wood chips gave increased impact of acidification and air quality (VOC and particles). The extra emissions of these substances are comparably low and can be decreased if excess heat from the mill is available.

Plastic pollution is a serious environmental problem. The fact that PHA-polymers are biodegradable, give positive environmental impacts that are not possible to measure with an E-LCA. Using PHA polymer products instead of fossil-based polymer-products, will have positive impact on marine, plant and animal life.

In this study there was too little information about fish feed production and fish farming for conducting an E-LCA. However, if Caldicellulosiruptor-biomass could avoid use of soya protein and fish meal it would have positive impact on marine environment, plant and animal life.

Feasibility

The summarized results indicate that the MultiBio concept is not feasible as described in this project. This is due to small production volumes. If the biohydrogen Caldicellulosiruptor-process can use a substrate with higher

source, indicate a possible future for such a cascading concept. If the infrastructure for hydrogen gas is developed in the near future, it might be possible to find collective use of smaller amounts of biohydrogen production.

Production of PHA-polymers is more flexible, and a large variation of organic waste can be used. This gives larger possibilities for process integration and larger production potentials. The large amount of organic waste at pulp and paper mills makes them suitable for hosting PHA-production. The positive environmental impact and the possibility to produce PHA to a reasonable cost, make production of PHA a feasible concept.

Communication and Dissemination

To communicate the MultiBio project we have used seven press releases, four in Swedish and three in English. See the titles below:

• (2018-06-15) I MultiBio blir brukens avfall biovätgas, bioplast och hållbart fiskfoder • (2019-03-12) Skapar fossilfria produkter av brukens restströmmar

• (2019-04-25) Pulp and paper mill waste becomes fish feed, energy and more • (2019-07-04) Alan tillverkar bioplast av avfall från massa- och pappersbruk • (2020-06-17) Bioplastics from pulp and paper side streams

• (2020-11-10) Biovätgas skapas ur reningsvatten på bruken

• (2020-12-02) Biohydrogen created from purification water at the mills

We have found 65 media notes about the project. Ten from regional media, 34 with a national coverage and 21 from international media. The media channels are mostly on the web (51pcs) but also one on YouTube, two on radio, four newspapers, six magazines and one book reference.

Table 6 Publications of MultiBio results on various platforms

# DATE Publication Publications Platform National/International

1 2018-06-21 Recycling Webb National

2 2018-06-25 Process Nordic Webb National

3 2018-06-25 Skog Supply Webb National

4 2018-06-25

Nyfiken på vad vi ska köra

våra bilar på Book National

5 2018-06-26 Papper och Massa Webb National

6 2018-11-01 Forum för Bioekonomi 2 Magazine National

7 2018-11-05 Skogsindustrierna Webb National

8 2018-11-05 Skogsindustrierna YouTube National

9 2018-12-13 NWT Webb Regional

10 2018-12-13 NWT Newspaper Regional

11 2019-02-01 KHaktuellt 1 Magazine Regional

12 2019-03-12 Industritorget Webb National

13 2019-03-13 NWT Newspaper Regional

14 2019-03-13 NWT Webb Regional

15 2019-03-22 Papper och massa Webb National

16 2019-03-28 Miljö & Utveckling Webb National

17 2019-04-02 Forum för Bioekonomi 1 Magazine National

18 2019-04-10 Recycling Webb National

19 2019-04-10 Packnet Webb National

20 2019-04-12 Packnews.se Webb National

21 2019-04-12 Plastfokus Webb International

22 2019-04-15 EmballageFokus Webb International

23 2019-04-15 _{Nordisk Bioplastförening} Webb National

25 2019-04-24 PaperAge Webb International

26 2019-04-25 Recycling Magazine Webb International

27 2019-04-25 PaperFIRST Webb International

28 2019-04-25 RecyclingPortal Webb International

29 2019-04-25

Pulp and paper mill

association Pakistan Webb International

30 2019-04-26 Pulpapernews Webb International

31 2019-04-26 Appita Webb International

32 2019-04-27 P4 Värmland Webb National

33 2019-04-27 P4 Värmland Radio National

34 2019-04-29 Packaging Europe Webb International

35 2019-05-10 Etiq & Pack Webb International

36 2019-05-23 aquaculturemagazine Webb International

37 2019-05-28 aqua culture brasil Webb International

38 2019-06-07 P4 Värmland Webb National

39 2019-06-07 P4 Värmland Radio National

40 2019-07-04 Papper och Massa Webb National

41 2019-07-08 Skogs Sverige Webb National

42 2019-10-01 Appita Magazine 4 Magazine International

43 2020-06-18 Pulpapernews Webb International

44 2020-06-18 Packthing Webb International

45 2020-06-23 The Bioeconomy Region Webb International

46 2020-10-19 B Magazine Webb International

47 2020-10-19 B Magazine Magazine International

48 2020-11-10 NWT Webb Regional

49 2020-11-10 Fryksdalsbygden Webb Regional

50 2020-11-10 Bioenergitidningen Webb National

51 2020-11-10 Värmlands Affärer Webb Regional

52 2020-11-10 elbilsnytt Webb National

53 2020-11-11 NWT Newspaper Regional

54 2020-11-11 Energinyheter Webb National

55 2020-11-11 Dagens Miljöteknik Webb National

56 2020-11-11 Papper och massa Webb National

57 2020-11-11 Renare Mark Webb National

58 2020-11-12 Skogs Sverige Webb National

59 2020-11-12 varim Webb National

60 2020-11-12 Miljö & Utveckling Webb National

61 2020-11-12 Recycling Webb National

62 2020-11-13 Fryksdalsbygden Newspaper Regional

63 2020-11-16 Process Nordic Webb National

64 2020-12-01 PaperFIRST Webb International

65 2020-12-02 Pulpapernews Webb International

One other dissemination result is Johanna Björkmalms licentiate dissertation “Design of kinetic models for assessment of critical aspects in bioprocess development: A case study of biohydrogen”.

Figure 14 Front-cover of the licentiate dissertation of Johanna Björkmalm

Five degree papers have been initiated; two at Karlstad University; one at Swedish University of Agricultural Sciences and two at Lund University. Three papers are ready:

• Philip Larsson “Tillverkning av polymerer från skogsindustriellt bioslam”, Karlstad University

• Erik Örnflo ”Design and simulation of biohydrogen & biopolymer production using byproduct streams from pulp and paper industries”, Lund University

• Fritiof Pröjts Erlandsson “A future introduction of biohydrogen to the Swedish energy system: Current developments and opportunities from the multi-level perspective”, Lund University

A detailed technical report has also been disseminated:

Werker, A. and Bengtsson, S. (2020). The production of biopolymers for bioplastics using pulp and paper mill wastewater and

residual fibre streams, Promiko AB Report 18-01-A, RISE, DiVA id: diva2:1506255, URN: urn:nbn:se:ri:diva-50940. The project has been presented on following seminar and conference events:

Table 7 List of events where MultiBio was presented

Presenter Date Event

Maria Sandberg 2018-04-23 Seminarieserie på Karlstads universitet

Karin Willquist 2018-09-18 Nya foder‐råvaror till fisk och kräftdjur för utveckling av hållbart vattenbruk

Maria Sandberg 2018-11-28 ÅF Vattenseminarie 2018 Maria Sandberg & Margareta

Sandström 2018-12-05 Sverige kraftsamlar för industriell symbios (SNIUS 5) Stefan Ivarsson 2020-03-26 Lansering av "Värmlands energi – och klimatstrategi" Sudhanshu Pawar 2020-10-13 Nordic Wood Biorefinery Conference 2020

The final seminar was made available to everyone to join via a digital platform and was recorded. The recording can be found here – Part 1: https://vimeo.com/486752553 and Part 2: https://vimeo.com/486753803

Conclusions

A fiber sludge containing, CTMP effluent process stream was considered to be a suitable residual stream for the proposed MultiBio concept. Due to safety risks the demo-scale experiments of biohydrogen gas technology were moved from Biorefinery demo plant (Örnsköldsvik) of 40 m3 _{capacity to ATEX classified pilot-scale facility with 0.4 m}3 capacity. Hence, bacterial biomass production that would be enough for the planned larger scale fish feed ingredient could not be produced. Instead, lab-scale experiments were performed with Caldicellulosiruptor cells as fish feed ingredient and these showed promising results as a protein-rich, sustainable fish feed ingredient. However, the production process of the dried bacterial biomass also highlighted the technical and potentially economic challenges with its production for a large-scale implementation. Alternatively, PHA biopolymer was used as fish food ingredient for experiments at Gårdsfisk AB with favourable results. Lab-scale experimental tests found that the mill surplus activated sludge from the mill wastewater treatment processes today would accumulate PHA to about 20 % of its organic weight. Mass balance evaluations based on realistically achievable expectations indicated a PHA biopolymer production potential of 3 600 tons of PHA per year from available organic residuals and combined for the two mills evaluated. The concept has a positive climate impact in comparison with current treatment and takes us in a positive direction to achieve 7 of 10 Swedish environmental goals. Through a detailed feasibility analysis two scenarios were suggested. The MultiBio cascade process can be implemented with further necessary development with good business potential and a positive effect on climate change. The biohydrogen technology needs further development before it can be implemented at industrial scale. A scenario with only biopolymer technology shows a larger business potential, even larger positive effect on climate change and with industrial implementation achievable in the shorter term. A

successful pilot demonstration of this biopolymer production scenario would help to support the development and integration of a win-win symbiosis with present day mill operations. Furthermore, MultiBio has attracted a lot of attention regionally and nationally but also internationally with a total of 65 media listings. Overall, MultiBio has successfully achieved its goals and objectives in exploring opportunities and challenges for renewable resources.

RISE Research Institutes of Sweden AB Energy and Circular Economy

Box 857, SE-501 15 BORÅS, Sweden Telephone: +46 10 516 50 00