- Email: [email protected]
Dewatering investigations on fungal biomass grown in thin stillage from a dry-mill corn ethanol plant
Biomass and Bioenergy 97 (2017) 65e69
Contents lists available at ScienceDirect
Biomass and Bioenergy journal homepage: http://www.elsevier.com/locate/biombioe
Short communication
Dewatering investigations on fungal biomass grown in thin stillage from a dry-mill corn ethanol plant Christopher R. Koza a, Glenn A. Norton b, J. (Hans) van Leeuwen b, * a b
Burns & McDonnell, 9400 Ward Parkway, Kansas City, MO 64114, USA Civil, Environmental, and Construction Engineering, Iowa State University, Ames, IA 50011, USA
a r t i c l e i n f o
a b s t r a c t
Article history: Received 15 June 2016 Received in revised form 6 October 2016 Accepted 13 December 2016
An innovative bioprocess utilizing thin stillage from a dry-grind corn ethanol plant was used to produce a useful filamentous fungus (Rhizopus microsporus var. oligosporus) in a pilot-scale bioreactor. The fungal process can improve the economics of corn ethanol production by producing an excellent food supplement for livestock or serving as a feedstock material for producing chitin, chitosan, and glucosamine. However, in order to be economically viable, effective and low-cost mechanical dewatering of the fungal biomass grown in thin stillage is required. In this study, dewatering tests were performed on fungal biomass using gravity and centrifugal sedimentation, gravity screening, a belt filter, a filter press, and centrifuge filtration in order to determine the most effective dewatering methods for this application. Utilizing a gravity-fed concave screen followed by a centrifuge filter proved to be the most effective dewatering approach and increased the screenable solids (i.e., larger than 20 mesh) content of the fungal biomass from the bioreactor from 1% to 30%. Achieving a solids content greater than 30% with mechanical dewatering is unlikely because of theoretical limits due to intracellular water. Nonetheless, this degree of dewatering greatly reduces thermal drying costs necessary to obtain a final product with a moisture content of 10%. © 2016 Elsevier Ltd. All rights reserved.
Keywords: Fungus Rhizopus oligosporus Fermentation Ethanol Dewatering Thin stillage
1. Introduction Between 2004 and 2014, the U.S. production of ethanol increased from 3.4 to 14.3 million gallons annually [1], with drymill plants accounting for over 80% of the ethanol produced [2]. In the dry-mill process, a low-value byproduct stream known as thin stillage is produced. Some of the thin stillage, which contains solubles and some residual suspended solids, is recycled back to the fermentation tanks. However, the amount of thin stillage that can be recycled is limited by its lactic acid, acetic acid, and glycerol concentrations, which build up in the fermentation tanks and inhibit ethanol production. The remaining thin stillage is evaporated to obtain syrup that can be used to produce distiller's dried grains with solubles (DDGS). Producing additional co-products would improve the economics of the ethanol industry, and cultivating the filamentous fungus (Rhizopus microsporus var. oligosporus, or simply
* Corresponding author. Department of Civil, Construction, and Environmental Engineering, 476 Town Engineering, Iowa State University, Ames, IA, 50011, USA. E-mail address: [email protected] (J. van Leeuwen). http://dx.doi.org/10.1016/j.biombioe.2016.12.011 0961-9534/© 2016 Elsevier Ltd. All rights reserved.
R. oligosporus) on thin stillage is one possible option [3e5]. Other fungal species have also been successfully grown in thin stillage [6,7]. The fungi consume many organic compounds in the thin stillage, including lactic acid, acetic acid, and glycerol [3,4]. In addition, a nutritional co-product having high concentrations of crude protein and essential amino acids for use as an animal feed is produced [8]. The fungi can also serve as a feedstock material for producing chitin, chitosan, and glucosamine. However, in order to be economically viable, extensive and cost-effective mechanical dewatering of the fungi grown in thin stillage in the bioreactor is required. Various mechanical dewatering approaches were tested for that purpose. The impact of the dewatering technologies on the nutrient content of the fungi was beyond the scope of this study. This work is not intended to present a thorough or comprehensive evaluation of the various dewatering technologies investigated. Rather, it is intended only to serve as exploratory tests to determine which of the specific methods and specific pieces of equipment tested yielded encouraging results, and to use that information to gain insights into suitable dewatering approaches for our continuing research on utilizing fungal biomass grown in thin stillage.
66
C.R. Koza et al. / Biomass and Bioenergy 97 (2017) 65e69
2. Experimental methods 2.1. Fungal cultivation Fungal biomass was grown in thin stillage in a 1600-L pilot-scale bioreactor containing 1500 L of thin stillage. Thin stillage collected from the Lincolnway Energy (Nevada, IA) plant was used as the growth medium. Spore suspensions of R. oligosporus were initially prepared and stored according to Ozsoy et al. [5]. Fungal inoculant for the reactor was then prepared by adding one 2-mL vial of spore suspension (containing 105 to 106 spores/mL) to each of eight 2-L flasks containing one liter of sterilized YM broth. Those seed cultures were incubated at 37 C for 24 h while shaking at 180 rpm. After the thin stillage in the bioreactor reached 37 C, it was inoculated with the seed cultures and the reactor was operated for 48 h in batch mode. Air was introduced through fine-bubble ceramic diffusers at a rate of 200e300 L/min in order to provide dissolved oxygen and promote mixing. The effluent from the bioreactor consisted of cultivated fungal biomass suspended in residual thin stillage. Unless otherwise noted, this reactor effluent was used for each of the solid-liquid separation techniques investigated. Prior to fungal cultivation, the thin stillage contained 3% total suspended solids. However, the concentration of the suspended solids that could be removed with a 20-mesh screen (hereafter referred to as “screenable solids”) was only 0.1% (w/w). After fungal cultivation, the concentration of total suspended solids in the reactor effluent remained at 3%, but the concentration of screenable solids was 1% (w/w) and consisted of fungal mycelia which grew in the form of pellets.
2.2. Biomass dewatering A variety of techniques for dewatering the effluent from the bioreactor were explored. Samples of dewatered fungal biomass were analyzed for moisture content according to Mitra et al. [6], while total solids, total suspended solids, and total dissolved solids were determined using standard procedures for wastewater [9]. Screenable solids were calculated after passing the fungal biomass over a 20-mesh screen.
2.2.2. Filtration The specific resistance to filtration (SRF) was tested according to Coackley and Jones [11]. This test provides insight into the dewaterability of solid-liquid slurries and has been extensively utilized for sludge dewaterability in the wastewater treatment industry. For this test, a Buchner funnel was used with Whatman 11-cm, Grade-4 disk filters having a pore size of 20e25 mm. The Buchner funnel was inserted into the top of a graduated cylinder that was modified by adding a side-arm, and a vacuum pump was attached to the sidearm to provide suction for the filtration. For each test, 200-mL samples of reactor effluent were vacuum-filtered at a negative pressure of 30 kPa. The cumulative volume of filtrate was plotted vs. time and the slope of the line was used to calculate the SRF value. After filtration, the mass of filter cake per unit volume of filtrate was determined by drying and weighing. The specific resistance to filtration (“r”, in m/kg) was then calculated using Eq (2), where “DP00 is the transmembrane pressure (Pa, or kg m1 s2), “A” is the filter area (m2), “b” is the slope of the line from the time vs. cumulative volume plot (s/m6), “m” is the dynamic viscosity of the filtrate (kg m1s1), and “w” is mass of cake per unit volume filtrate (kg/m3).
. r ¼ 2DPA2 b mw
Gravity screening was investigated using a 20-mesh, inclined, variable-slope screen that was 1.2 m in length (see Fig. 1). As the bioreactor effluent was discharged over the screen, most of the dewatering occurred at the top of the screen, which was inclined at 65 . The biomass accumulated at the bottom of the screen and was scraped into a hopper, while the filtrate was collected underneath the screen and discharged through a flow meter. The volume of filtrate and the mass of screened biomass were measured. Tests with a pilot-scale belt filter were conducted by Siemens Water Technologies. An initial test involved using reactor effluent that had been passed over a 20-mesh screen to increase the screenable solids content from 1% to 15%. However, the solids content was too high and could not be fed into the belt filter. Two tests were subsequently performed using the bioreactor effluent
2.2.1. Sedimentation Laboratory tests were conducted to evaluate dewatering by sedimentation. Settling characteristics of the suspended solids in the bioreactor effluent were assessed by determining the sludge volume index (SVI) [10]. The SVI value, reported in mL/g, is the volume (mL) occupied by 1 g of suspension after settling for 30 min. It is defined according to Eq (1), where “SV3000 is the volume of settled sludge after 30 min in a 1-L graduated cylinder (mL/L), and “x” is the suspended solids concentration of the bioreactor effluent (g/L).
SVI ¼ SV30 =x
(1)
Subsequent tests were performed using a laboratory centrifuge to determine the dewatering potential of the bioreactor effluent at various g-forces. Centrifuge bottles were filled with 250 ml of the reactor effluent and run at 2000e5000 g's for 10 min. Sludge volumes were calculated as the percent of settled fungal biomass after centrifuging. In addition to using a laboratory centrifuge, 200-L samples of reactor effluent were run through a pilot-scale horizontal decanter centrifuge (Centrisys, Kenosha, WI, USA). The speed of the decanter centrifuge was increased from 1000 g's to its upper limit of 3500 g's while adjusting the screw conveyer speed.
(2)
Fig. 1. Simple gravity screen used for filtration tests.
C.R. Koza et al. / Biomass and Bioenergy 97 (2017) 65e69
containing 3% suspended solids. One test each was performed with and without the addition of a cationic polymer to the reactor effluent prior to passing the effluent through the belt filter. A polymer dosing of 200 mg/L of thin stillage was used, which corresponds to a polymer dosing of 6.7 g/kg of suspended solids (dry basis). Tests with a bench-top filter press were also performed by Siemens Water Technologies. Samples of the reactor effluent containing 3% suspended solids were placed into the sample reservoir and pressurized with compressed air. An initial feed pressure of 25 psig was used and then increased to 100 and 225 psig. The filtration cycle ended when the filtrate flow reached 0.01 gal/min per ft2 of filtration area. Centrifuge filtration was also investigated. However, since economical small-scale simulation equipment was not available, a domestic washing machine was used to assess that approach. The machine's 26-cm drum radius spun at 800 rpm and produced a force of 200 g's, which is equivalent to those found in industrial centrifuge filters. The g-force was calculated using Eq (3), where RCF (expressed as multiple g's) is the relative centrifugal force, “g” is the earth's gravitational acceleration, “r” is the rotational radius of the centrifuge filter, and “u” is the angular velocity in radians per unit time.
. RCF ¼ r u2 g
(3)
This relationship may be approximated by Eq (4), where “rmm” is the rotational radius in millimeters (mm), and “NRPM” is the rotational speed in rpm.
RCF ¼ 1:12 106 rmm N 2 rpm
(4)
Samples of reactor effluent that had been pre-screened using the apparatus in Fig. 1 to produce wet fungal biomass containing 15% screenable solids were transferred to woven polypropylene bags, placed in the washing machine, and “processed” using the washer's spin cycle. The polyethylene bags retained the fungal biomass while liquid was removed as a result of spinning in the washer's drum during the spin cycle. For these dewatering tests, biomass loading rates of 8, 9, and 10 kg (dry basis) per square meter of filter area were tested, along with filtration times of 5, 10, 15, 20, and 25 min. Samples of dewatered biomass from the washing machine were taken at 5-min intervals to track moisture content. For examining cake thickness (indicating compressibility of the cake) between run times of 5 and 25 min, sample loadings of 10, 15, and 20 kg/m2 of filter area were used. Single tests were performed under each set of conditions using the washing machine as a centrifuge filter. 3. Results and discussion 3.1. Sedimentation Results from the SVI tests indicated that gravity sedimentation was ineffective at dewatering the effluent from the bioreactor. No settling was achieved in multiple tests, which shows that the reactor effluent is beyond hindered settling whereby interparticle forces hinder settling of neighboring particles [10]. The specific gravity of the fungal biomass is 1.03, and the small density differential between the fungi and water resulted in low settling velocities. The laboratory centrifuge operating at 2000 to 5000 g's and the pilot-scale continuous decanter centrifuge operating at 1000e3500 g's also proved to be ineffective at dewatering the reactor effluent. At 3500 g's in the decanter centrifuge, there was a
67
breakdown of the fungal fibrous structure in the fungal pellets and the high g-forces turned the fungal mycelia into a gelatinous paste.
3.2. Filtration Results of the tests on specific resistance to filtration (SRF) showed that the effluent from the fungal bioreactor had an SRV value of 2.0 1013, which is substantially less than bacterial sludges from a wastewater treatment plant. The lower SRF values relative to some wastewater sludges, coupled with readily available filtration equipment for sludge dewatering, warranted the investigation of filtration as a dewatering approach. During gravity screening tests, no visible loss of fungal biomass through the screen was observed. The filamentous nature of the fungi formed an interlocking matrix of fungal mycelia (which grew in the form of pellets) and prevented the biomass from passing through the screen while allowing much of the water to pass through. This thickened the biomass from 1% to 15% (w/w) screenable solids. Approximately 93 kg of water was removed for every kilogram of fungal biomass. This degree of dewatering is quite useful and only requires power for a scraping mechanism, making it a cost-effective approach for removing the bulk (>90%) of the water from the fungal biomass. In addition, with more than a ten-fold reduction in process volume, the size of subsequent dewatering equipment can be significantly reduced. Although gravity screening is a good first step, additional mechanical dewatering is needed in order for the overall drying process to be economical. It should be noted that the proper mesh size for dewatering the fungal biomass will depend on the physical nature of the fungi, including the size of the fungal pellets. These can vary considerably depending on the fungal species used and numerous variables in the growth environment. One patent involving growing fungi in thin stillage states that the pellets grow to 3e5 mm in diameter [12], while another patent states that dewatering fungal pellets might involve using screens with openings ranging from 1 to 8 mm [13]. There is no general rule on the mesh size that will work best, and it may be necessary to make that determination on a case by case basis. As noted earlier, the initial test with the pilot-scale belt filter used reactor effluent that had been passed over a 20-mesh sieve to provide fungal biomass containing 15% screenable solids, but the solids content was too high and could not be fed into the belt filter. When effluent (containing 3% suspended solids) directly from the bioreactor was used, the effluent could be effectively filtered with the belt filter, but only when adding 200 mg of cationic polymer per liter of thin stillage. Results indicated that obtaining a biomass cake with up to 25% solids can be obtained. However, this approach was ultimately deemed undesirable in view of the polymer costs and concerns about adding chemicals for a product intended to be used as animal feed. Tests using a bench-scale filter press indicated that the effluent from the bioreactor is not amenable to pressure filtration. During testing, a 25-mm thick cake layer formed and very little water was removed because solids in the reactor effluent rapidly fouled the membrane. Also, release of the biomass cake from the filtration unit was difficult. The filter press had large working pressures of 100e225 psig, which led to excessive cake compression. This lowered permeability, increased specific resistance to filtration, and lowered filtration efficiency. Biological sludges are known to exhibit high compressibility, low permeability, and high specific resistance to filtration at moderate to high pressure.
68
C.R. Koza et al. / Biomass and Bioenergy 97 (2017) 65e69
31
30
Solids Content (%)
29
28
27
8 kg dry solids / m^2 filter area 9 kg dry solids / m^2 filter area
26
10 kg dry solids / m^2 filter area
25
24 0
5
10
15
20
25
30
FiltraƟon Time (min) Fig. 2. Dewatering of fungal biomass (15% initial solids content) using a filter centrifuge.
Table 1 Compressibility of the filter cake from the filter centrifuge. Fungal Biomass Loading (kg/m2 filter area)
Initial Cake Thickness (mm)
Final Cake Thickness (cm)
Change in Cake Thickness (%)
10 15 20
38 42 51
27 30 38
29 29 26
3.3. Filter centrifugation The feed material for the filter centrifuge consisted of reactor effluent that had been dewatered with a 20-mesh screen to provide a solids content of 15%. Results of the tests using a centrifuge filter are shown in Fig. 2. The centrifuge filter was able to increase the solids content from 15% to nearly 30%. Attaining a solids content of more than about 30% using mechanical dewatering is unlikely due to the presence of intracellular water, which cannot be effectively removed through mechanical means. Time played a critical role in the dewatering efficiency, and the solids content of the filter cake plateaued after about 20 min. The filter centrifuge used a relatively low g-force (200 g's) compared to sedimentation centrifuges. Also, these results were achieved at lower pressures (10e20 psi) relative to other equipment used in this study. Low pressure and low g-forces were key factors affecting the success of this dewatering approach because it allowed the fungal biomass to retain its fibrous and porous structure while still providing a sufficient driving force to remove much of the water. As shown in Table 1, there was only minimal cake compression. As a result, there was no significant increase in specific resistance to filtration and a relatively high solids content was achieved.
4. Summary and conclusions Gravity screening coupled with a filter centrifuge appears to be a viable option for dewatering fungal biomass grown in thin stillage. About 96% of the free liquid can be removed with that approach, producing a filter cake with about 30% solids over a short period of time. Producing a biomass cake with more than about 30% solids is unlikely when using mechanical dewatering approaches because of
theoretical limits due to intracellular water. Ultimately, additional liquid will need to be removed by thermal drying to produce fungal biomass with 90% solids. Nonetheless, the rapid and inexpensive removal of most of the water using gravity screening and a centrifuge filter will substantially reduce overall drying costs. Acknowledgements This research was supported through a grant (#09-05) from the Iowa Energy Center. We thank Norm Olson, Linda Hintch, Dr. Sipho Ndlela and other Iowa Energy Center staff members for their support of the project. Additionally, we would like to thank the staff at Lincolnway Energy for providing thin stillage samples and operating space for filtration tests. We also thank Daniel Erickson for assisting with some of the tests. References [1] Renewable Fuels Association, Industry statistics. http://www.ethanolrfa.org/ resources/industry/statistics (Accessed 11 January 2016). [2] U.S. Department of Energy, Alternative Fuels Data Center, Ethanol Production and Distribution, http://www.afdc.energy.gov/fuels/ethanol_production.html, (Accessed 11 January 2016). [3] M.L. Rasmussen, S.K. Khanal, A.L. Pometto III, J. (Hans) van Leeuwen, Water reclamation and value-added animal feed from corn-ethanol stillage by fungal processing, Bioresour. Technol. 151 (2014) 284e290. [4] J. (Hans) van Leeuwen, M.L. Rasmussen, S. Sankaran, C.R. Koza, D.T. Erickson, D. Mitra, B. Jin, Fungal treatment of crop processing wastewaters with valueadded co-products, in: K. Gopalakrishnan, J. van Leeuwen, R.C. Brown (Eds.), Sustainable Bioenergy and Byproducts, Springer-Verlag, London, 2012, pp. 13e44. [5] H.D. Ozsoy, H. Kumbur, B. Saha, J. van Leeuwen, Use of Rhizopus oligosporus produced from food processing wastewater as a biosorbent for Cu (II) ions removal from the aqueous solutions, Bioresour. Technol. 99 (2008) 4943e4948. [6] D. Mitra, M.L. Rasmussen, P. Chand, V.R. Chintareddy, L. Yao, D. Grewell, J.G. Verkade, T. Wang, J.(H.) van Leeuwen, Value-added oil and animal feed
C.R. Koza et al. / Biomass and Bioenergy 97 (2017) 65e69 production from corn-ethanol stillage using the oleaginous fungus Mucor circinelloides, Bioresour. Technol. 107 (2012) 368e375. [7] J.A. Ferreira, P.R. Lennartsson, M.J. Taherzadeh, Production of ethanol and biomass from thin stillage using food-grade Zygomycetes and Ascomycetes filamentous fungi, Energies 7 (2014) 3872e3885. [8] D. van Sambeek, T.E. Weber, B.J. Kerr, J. van Leeuwen, N. Gabler, Evaluation of Rhizopus oligosporus in nursery pig diets on growth performance and nutrient digestibility, J. Anim. Sci. 90 (Suppl. 2) (2012) 129. [9] American Public Health Asociation, American Water Works Association, Water Environment Federation, twenty-first ed., Standard Methods for the Examination of Water and Wastewater, APHA/AWWA/WEF, Washington DC, 2005.
69
[10] G. Tchobanoglous, F. Burton, H.D. Stensel, Wastewater Engineering: Treatment and Reuse, fourth ed., McGraw-Hill, New York, 2003. [11] P. Coackley, B.R.S. Jones B.R.S., Vacuum sludge filtration: interpretation of results by the concept of specific resistance, Sew. Ind. Wastes 28 (8) (1956) 963e976. [12] J. van Leeuwen, S.K. Khanal, A.L. Pometto, Purification of thin stillage from dry-grind corn milling with fungi, U. S. Pat. 9079786 Issued July 14, 2015, http://www.freepatentsonline.com/9079786.pdf. [13] J. van Leeuwen, S.K. Khanal, A.L. Pometto, M.L. Rasmussen, D. Mitra, Fungi cultivation on alcohol fermentation stillage for useful products and energy savings, U. S. Pat. 8481295 Issued July 9, 2013, https://docs.google.com/ viewer?url=patentimages.storage.googleapis.com/pdfs/US8481295.pdf.
Contents lists available at ScienceDirect
Biomass and Bioenergy journal homepage: http://www.elsevier.com/locate/biombioe
Short communication
Dewatering investigations on fungal biomass grown in thin stillage from a dry-mill corn ethanol plant Christopher R. Koza a, Glenn A. Norton b, J. (Hans) van Leeuwen b, * a b
Burns & McDonnell, 9400 Ward Parkway, Kansas City, MO 64114, USA Civil, Environmental, and Construction Engineering, Iowa State University, Ames, IA 50011, USA
a r t i c l e i n f o
a b s t r a c t
Article history: Received 15 June 2016 Received in revised form 6 October 2016 Accepted 13 December 2016
An innovative bioprocess utilizing thin stillage from a dry-grind corn ethanol plant was used to produce a useful filamentous fungus (Rhizopus microsporus var. oligosporus) in a pilot-scale bioreactor. The fungal process can improve the economics of corn ethanol production by producing an excellent food supplement for livestock or serving as a feedstock material for producing chitin, chitosan, and glucosamine. However, in order to be economically viable, effective and low-cost mechanical dewatering of the fungal biomass grown in thin stillage is required. In this study, dewatering tests were performed on fungal biomass using gravity and centrifugal sedimentation, gravity screening, a belt filter, a filter press, and centrifuge filtration in order to determine the most effective dewatering methods for this application. Utilizing a gravity-fed concave screen followed by a centrifuge filter proved to be the most effective dewatering approach and increased the screenable solids (i.e., larger than 20 mesh) content of the fungal biomass from the bioreactor from 1% to 30%. Achieving a solids content greater than 30% with mechanical dewatering is unlikely because of theoretical limits due to intracellular water. Nonetheless, this degree of dewatering greatly reduces thermal drying costs necessary to obtain a final product with a moisture content of 10%. © 2016 Elsevier Ltd. All rights reserved.
Keywords: Fungus Rhizopus oligosporus Fermentation Ethanol Dewatering Thin stillage
1. Introduction Between 2004 and 2014, the U.S. production of ethanol increased from 3.4 to 14.3 million gallons annually [1], with drymill plants accounting for over 80% of the ethanol produced [2]. In the dry-mill process, a low-value byproduct stream known as thin stillage is produced. Some of the thin stillage, which contains solubles and some residual suspended solids, is recycled back to the fermentation tanks. However, the amount of thin stillage that can be recycled is limited by its lactic acid, acetic acid, and glycerol concentrations, which build up in the fermentation tanks and inhibit ethanol production. The remaining thin stillage is evaporated to obtain syrup that can be used to produce distiller's dried grains with solubles (DDGS). Producing additional co-products would improve the economics of the ethanol industry, and cultivating the filamentous fungus (Rhizopus microsporus var. oligosporus, or simply
* Corresponding author. Department of Civil, Construction, and Environmental Engineering, 476 Town Engineering, Iowa State University, Ames, IA, 50011, USA. E-mail address: [email protected] (J. van Leeuwen). http://dx.doi.org/10.1016/j.biombioe.2016.12.011 0961-9534/© 2016 Elsevier Ltd. All rights reserved.
R. oligosporus) on thin stillage is one possible option [3e5]. Other fungal species have also been successfully grown in thin stillage [6,7]. The fungi consume many organic compounds in the thin stillage, including lactic acid, acetic acid, and glycerol [3,4]. In addition, a nutritional co-product having high concentrations of crude protein and essential amino acids for use as an animal feed is produced [8]. The fungi can also serve as a feedstock material for producing chitin, chitosan, and glucosamine. However, in order to be economically viable, extensive and cost-effective mechanical dewatering of the fungi grown in thin stillage in the bioreactor is required. Various mechanical dewatering approaches were tested for that purpose. The impact of the dewatering technologies on the nutrient content of the fungi was beyond the scope of this study. This work is not intended to present a thorough or comprehensive evaluation of the various dewatering technologies investigated. Rather, it is intended only to serve as exploratory tests to determine which of the specific methods and specific pieces of equipment tested yielded encouraging results, and to use that information to gain insights into suitable dewatering approaches for our continuing research on utilizing fungal biomass grown in thin stillage.
66
C.R. Koza et al. / Biomass and Bioenergy 97 (2017) 65e69
2. Experimental methods 2.1. Fungal cultivation Fungal biomass was grown in thin stillage in a 1600-L pilot-scale bioreactor containing 1500 L of thin stillage. Thin stillage collected from the Lincolnway Energy (Nevada, IA) plant was used as the growth medium. Spore suspensions of R. oligosporus were initially prepared and stored according to Ozsoy et al. [5]. Fungal inoculant for the reactor was then prepared by adding one 2-mL vial of spore suspension (containing 105 to 106 spores/mL) to each of eight 2-L flasks containing one liter of sterilized YM broth. Those seed cultures were incubated at 37 C for 24 h while shaking at 180 rpm. After the thin stillage in the bioreactor reached 37 C, it was inoculated with the seed cultures and the reactor was operated for 48 h in batch mode. Air was introduced through fine-bubble ceramic diffusers at a rate of 200e300 L/min in order to provide dissolved oxygen and promote mixing. The effluent from the bioreactor consisted of cultivated fungal biomass suspended in residual thin stillage. Unless otherwise noted, this reactor effluent was used for each of the solid-liquid separation techniques investigated. Prior to fungal cultivation, the thin stillage contained 3% total suspended solids. However, the concentration of the suspended solids that could be removed with a 20-mesh screen (hereafter referred to as “screenable solids”) was only 0.1% (w/w). After fungal cultivation, the concentration of total suspended solids in the reactor effluent remained at 3%, but the concentration of screenable solids was 1% (w/w) and consisted of fungal mycelia which grew in the form of pellets.
2.2. Biomass dewatering A variety of techniques for dewatering the effluent from the bioreactor were explored. Samples of dewatered fungal biomass were analyzed for moisture content according to Mitra et al. [6], while total solids, total suspended solids, and total dissolved solids were determined using standard procedures for wastewater [9]. Screenable solids were calculated after passing the fungal biomass over a 20-mesh screen.
2.2.2. Filtration The specific resistance to filtration (SRF) was tested according to Coackley and Jones [11]. This test provides insight into the dewaterability of solid-liquid slurries and has been extensively utilized for sludge dewaterability in the wastewater treatment industry. For this test, a Buchner funnel was used with Whatman 11-cm, Grade-4 disk filters having a pore size of 20e25 mm. The Buchner funnel was inserted into the top of a graduated cylinder that was modified by adding a side-arm, and a vacuum pump was attached to the sidearm to provide suction for the filtration. For each test, 200-mL samples of reactor effluent were vacuum-filtered at a negative pressure of 30 kPa. The cumulative volume of filtrate was plotted vs. time and the slope of the line was used to calculate the SRF value. After filtration, the mass of filter cake per unit volume of filtrate was determined by drying and weighing. The specific resistance to filtration (“r”, in m/kg) was then calculated using Eq (2), where “DP00 is the transmembrane pressure (Pa, or kg m1 s2), “A” is the filter area (m2), “b” is the slope of the line from the time vs. cumulative volume plot (s/m6), “m” is the dynamic viscosity of the filtrate (kg m1s1), and “w” is mass of cake per unit volume filtrate (kg/m3).
. r ¼ 2DPA2 b mw
Gravity screening was investigated using a 20-mesh, inclined, variable-slope screen that was 1.2 m in length (see Fig. 1). As the bioreactor effluent was discharged over the screen, most of the dewatering occurred at the top of the screen, which was inclined at 65 . The biomass accumulated at the bottom of the screen and was scraped into a hopper, while the filtrate was collected underneath the screen and discharged through a flow meter. The volume of filtrate and the mass of screened biomass were measured. Tests with a pilot-scale belt filter were conducted by Siemens Water Technologies. An initial test involved using reactor effluent that had been passed over a 20-mesh screen to increase the screenable solids content from 1% to 15%. However, the solids content was too high and could not be fed into the belt filter. Two tests were subsequently performed using the bioreactor effluent
2.2.1. Sedimentation Laboratory tests were conducted to evaluate dewatering by sedimentation. Settling characteristics of the suspended solids in the bioreactor effluent were assessed by determining the sludge volume index (SVI) [10]. The SVI value, reported in mL/g, is the volume (mL) occupied by 1 g of suspension after settling for 30 min. It is defined according to Eq (1), where “SV3000 is the volume of settled sludge after 30 min in a 1-L graduated cylinder (mL/L), and “x” is the suspended solids concentration of the bioreactor effluent (g/L).
SVI ¼ SV30 =x
(1)
Subsequent tests were performed using a laboratory centrifuge to determine the dewatering potential of the bioreactor effluent at various g-forces. Centrifuge bottles were filled with 250 ml of the reactor effluent and run at 2000e5000 g's for 10 min. Sludge volumes were calculated as the percent of settled fungal biomass after centrifuging. In addition to using a laboratory centrifuge, 200-L samples of reactor effluent were run through a pilot-scale horizontal decanter centrifuge (Centrisys, Kenosha, WI, USA). The speed of the decanter centrifuge was increased from 1000 g's to its upper limit of 3500 g's while adjusting the screw conveyer speed.
(2)
Fig. 1. Simple gravity screen used for filtration tests.
C.R. Koza et al. / Biomass and Bioenergy 97 (2017) 65e69
containing 3% suspended solids. One test each was performed with and without the addition of a cationic polymer to the reactor effluent prior to passing the effluent through the belt filter. A polymer dosing of 200 mg/L of thin stillage was used, which corresponds to a polymer dosing of 6.7 g/kg of suspended solids (dry basis). Tests with a bench-top filter press were also performed by Siemens Water Technologies. Samples of the reactor effluent containing 3% suspended solids were placed into the sample reservoir and pressurized with compressed air. An initial feed pressure of 25 psig was used and then increased to 100 and 225 psig. The filtration cycle ended when the filtrate flow reached 0.01 gal/min per ft2 of filtration area. Centrifuge filtration was also investigated. However, since economical small-scale simulation equipment was not available, a domestic washing machine was used to assess that approach. The machine's 26-cm drum radius spun at 800 rpm and produced a force of 200 g's, which is equivalent to those found in industrial centrifuge filters. The g-force was calculated using Eq (3), where RCF (expressed as multiple g's) is the relative centrifugal force, “g” is the earth's gravitational acceleration, “r” is the rotational radius of the centrifuge filter, and “u” is the angular velocity in radians per unit time.
. RCF ¼ r u2 g
(3)
This relationship may be approximated by Eq (4), where “rmm” is the rotational radius in millimeters (mm), and “NRPM” is the rotational speed in rpm.
RCF ¼ 1:12 106 rmm N 2 rpm
(4)
Samples of reactor effluent that had been pre-screened using the apparatus in Fig. 1 to produce wet fungal biomass containing 15% screenable solids were transferred to woven polypropylene bags, placed in the washing machine, and “processed” using the washer's spin cycle. The polyethylene bags retained the fungal biomass while liquid was removed as a result of spinning in the washer's drum during the spin cycle. For these dewatering tests, biomass loading rates of 8, 9, and 10 kg (dry basis) per square meter of filter area were tested, along with filtration times of 5, 10, 15, 20, and 25 min. Samples of dewatered biomass from the washing machine were taken at 5-min intervals to track moisture content. For examining cake thickness (indicating compressibility of the cake) between run times of 5 and 25 min, sample loadings of 10, 15, and 20 kg/m2 of filter area were used. Single tests were performed under each set of conditions using the washing machine as a centrifuge filter. 3. Results and discussion 3.1. Sedimentation Results from the SVI tests indicated that gravity sedimentation was ineffective at dewatering the effluent from the bioreactor. No settling was achieved in multiple tests, which shows that the reactor effluent is beyond hindered settling whereby interparticle forces hinder settling of neighboring particles [10]. The specific gravity of the fungal biomass is 1.03, and the small density differential between the fungi and water resulted in low settling velocities. The laboratory centrifuge operating at 2000 to 5000 g's and the pilot-scale continuous decanter centrifuge operating at 1000e3500 g's also proved to be ineffective at dewatering the reactor effluent. At 3500 g's in the decanter centrifuge, there was a
67
breakdown of the fungal fibrous structure in the fungal pellets and the high g-forces turned the fungal mycelia into a gelatinous paste.
3.2. Filtration Results of the tests on specific resistance to filtration (SRF) showed that the effluent from the fungal bioreactor had an SRV value of 2.0 1013, which is substantially less than bacterial sludges from a wastewater treatment plant. The lower SRF values relative to some wastewater sludges, coupled with readily available filtration equipment for sludge dewatering, warranted the investigation of filtration as a dewatering approach. During gravity screening tests, no visible loss of fungal biomass through the screen was observed. The filamentous nature of the fungi formed an interlocking matrix of fungal mycelia (which grew in the form of pellets) and prevented the biomass from passing through the screen while allowing much of the water to pass through. This thickened the biomass from 1% to 15% (w/w) screenable solids. Approximately 93 kg of water was removed for every kilogram of fungal biomass. This degree of dewatering is quite useful and only requires power for a scraping mechanism, making it a cost-effective approach for removing the bulk (>90%) of the water from the fungal biomass. In addition, with more than a ten-fold reduction in process volume, the size of subsequent dewatering equipment can be significantly reduced. Although gravity screening is a good first step, additional mechanical dewatering is needed in order for the overall drying process to be economical. It should be noted that the proper mesh size for dewatering the fungal biomass will depend on the physical nature of the fungi, including the size of the fungal pellets. These can vary considerably depending on the fungal species used and numerous variables in the growth environment. One patent involving growing fungi in thin stillage states that the pellets grow to 3e5 mm in diameter [12], while another patent states that dewatering fungal pellets might involve using screens with openings ranging from 1 to 8 mm [13]. There is no general rule on the mesh size that will work best, and it may be necessary to make that determination on a case by case basis. As noted earlier, the initial test with the pilot-scale belt filter used reactor effluent that had been passed over a 20-mesh sieve to provide fungal biomass containing 15% screenable solids, but the solids content was too high and could not be fed into the belt filter. When effluent (containing 3% suspended solids) directly from the bioreactor was used, the effluent could be effectively filtered with the belt filter, but only when adding 200 mg of cationic polymer per liter of thin stillage. Results indicated that obtaining a biomass cake with up to 25% solids can be obtained. However, this approach was ultimately deemed undesirable in view of the polymer costs and concerns about adding chemicals for a product intended to be used as animal feed. Tests using a bench-scale filter press indicated that the effluent from the bioreactor is not amenable to pressure filtration. During testing, a 25-mm thick cake layer formed and very little water was removed because solids in the reactor effluent rapidly fouled the membrane. Also, release of the biomass cake from the filtration unit was difficult. The filter press had large working pressures of 100e225 psig, which led to excessive cake compression. This lowered permeability, increased specific resistance to filtration, and lowered filtration efficiency. Biological sludges are known to exhibit high compressibility, low permeability, and high specific resistance to filtration at moderate to high pressure.
68
C.R. Koza et al. / Biomass and Bioenergy 97 (2017) 65e69
31
30
Solids Content (%)
29
28
27
8 kg dry solids / m^2 filter area 9 kg dry solids / m^2 filter area
26
10 kg dry solids / m^2 filter area
25
24 0
5
10
15
20
25
30
FiltraƟon Time (min) Fig. 2. Dewatering of fungal biomass (15% initial solids content) using a filter centrifuge.
Table 1 Compressibility of the filter cake from the filter centrifuge. Fungal Biomass Loading (kg/m2 filter area)
Initial Cake Thickness (mm)
Final Cake Thickness (cm)
Change in Cake Thickness (%)
10 15 20
38 42 51
27 30 38
29 29 26
3.3. Filter centrifugation The feed material for the filter centrifuge consisted of reactor effluent that had been dewatered with a 20-mesh screen to provide a solids content of 15%. Results of the tests using a centrifuge filter are shown in Fig. 2. The centrifuge filter was able to increase the solids content from 15% to nearly 30%. Attaining a solids content of more than about 30% using mechanical dewatering is unlikely due to the presence of intracellular water, which cannot be effectively removed through mechanical means. Time played a critical role in the dewatering efficiency, and the solids content of the filter cake plateaued after about 20 min. The filter centrifuge used a relatively low g-force (200 g's) compared to sedimentation centrifuges. Also, these results were achieved at lower pressures (10e20 psi) relative to other equipment used in this study. Low pressure and low g-forces were key factors affecting the success of this dewatering approach because it allowed the fungal biomass to retain its fibrous and porous structure while still providing a sufficient driving force to remove much of the water. As shown in Table 1, there was only minimal cake compression. As a result, there was no significant increase in specific resistance to filtration and a relatively high solids content was achieved.
4. Summary and conclusions Gravity screening coupled with a filter centrifuge appears to be a viable option for dewatering fungal biomass grown in thin stillage. About 96% of the free liquid can be removed with that approach, producing a filter cake with about 30% solids over a short period of time. Producing a biomass cake with more than about 30% solids is unlikely when using mechanical dewatering approaches because of
theoretical limits due to intracellular water. Ultimately, additional liquid will need to be removed by thermal drying to produce fungal biomass with 90% solids. Nonetheless, the rapid and inexpensive removal of most of the water using gravity screening and a centrifuge filter will substantially reduce overall drying costs. Acknowledgements This research was supported through a grant (#09-05) from the Iowa Energy Center. We thank Norm Olson, Linda Hintch, Dr. Sipho Ndlela and other Iowa Energy Center staff members for their support of the project. Additionally, we would like to thank the staff at Lincolnway Energy for providing thin stillage samples and operating space for filtration tests. We also thank Daniel Erickson for assisting with some of the tests. References [1] Renewable Fuels Association, Industry statistics. http://www.ethanolrfa.org/ resources/industry/statistics (Accessed 11 January 2016). [2] U.S. Department of Energy, Alternative Fuels Data Center, Ethanol Production and Distribution, http://www.afdc.energy.gov/fuels/ethanol_production.html, (Accessed 11 January 2016). [3] M.L. Rasmussen, S.K. Khanal, A.L. Pometto III, J. (Hans) van Leeuwen, Water reclamation and value-added animal feed from corn-ethanol stillage by fungal processing, Bioresour. Technol. 151 (2014) 284e290. [4] J. (Hans) van Leeuwen, M.L. Rasmussen, S. Sankaran, C.R. Koza, D.T. Erickson, D. Mitra, B. Jin, Fungal treatment of crop processing wastewaters with valueadded co-products, in: K. Gopalakrishnan, J. van Leeuwen, R.C. Brown (Eds.), Sustainable Bioenergy and Byproducts, Springer-Verlag, London, 2012, pp. 13e44. [5] H.D. Ozsoy, H. Kumbur, B. Saha, J. van Leeuwen, Use of Rhizopus oligosporus produced from food processing wastewater as a biosorbent for Cu (II) ions removal from the aqueous solutions, Bioresour. Technol. 99 (2008) 4943e4948. [6] D. Mitra, M.L. Rasmussen, P. Chand, V.R. Chintareddy, L. Yao, D. Grewell, J.G. Verkade, T. Wang, J.(H.) van Leeuwen, Value-added oil and animal feed
C.R. Koza et al. / Biomass and Bioenergy 97 (2017) 65e69 production from corn-ethanol stillage using the oleaginous fungus Mucor circinelloides, Bioresour. Technol. 107 (2012) 368e375. [7] J.A. Ferreira, P.R. Lennartsson, M.J. Taherzadeh, Production of ethanol and biomass from thin stillage using food-grade Zygomycetes and Ascomycetes filamentous fungi, Energies 7 (2014) 3872e3885. [8] D. van Sambeek, T.E. Weber, B.J. Kerr, J. van Leeuwen, N. Gabler, Evaluation of Rhizopus oligosporus in nursery pig diets on growth performance and nutrient digestibility, J. Anim. Sci. 90 (Suppl. 2) (2012) 129. [9] American Public Health Asociation, American Water Works Association, Water Environment Federation, twenty-first ed., Standard Methods for the Examination of Water and Wastewater, APHA/AWWA/WEF, Washington DC, 2005.
69
[10] G. Tchobanoglous, F. Burton, H.D. Stensel, Wastewater Engineering: Treatment and Reuse, fourth ed., McGraw-Hill, New York, 2003. [11] P. Coackley, B.R.S. Jones B.R.S., Vacuum sludge filtration: interpretation of results by the concept of specific resistance, Sew. Ind. Wastes 28 (8) (1956) 963e976. [12] J. van Leeuwen, S.K. Khanal, A.L. Pometto, Purification of thin stillage from dry-grind corn milling with fungi, U. S. Pat. 9079786 Issued July 14, 2015, http://www.freepatentsonline.com/9079786.pdf. [13] J. van Leeuwen, S.K. Khanal, A.L. Pometto, M.L. Rasmussen, D. Mitra, Fungi cultivation on alcohol fermentation stillage for useful products and energy savings, U. S. Pat. 8481295 Issued July 9, 2013, https://docs.google.com/ viewer?url=patentimages.storage.googleapis.com/pdfs/US8481295.pdf.