Page 1
Design and Construction of a Lighting System to Illuminate a
Photobioreactor
A thesis presented to
the faculty of
the Russ College of Engineering and Technology
of Ohio University
In partial fulfillment
of the requirements for the degree
Masters of Science
Kyle J. Sink
November 2011
© 2011 Kyle J. Sink. All Right Reserved.
Page 2
2
This thesis titled
Design and Construction of a Lighting System to Illuminate a
Photobioreactor
by
KYLE J. SINK
has been approved for
the Department of Mechanical Engineering
and the Russ College of Engineering and Technology by
Greg Kremer
Associate Professor of Mechanical Engineering
Dennis Irwin
Dean, Russ College of Engineering and Technology
Page 3
3
ABSTRACT
SINK, KYLE J., M.S., November 2011, Mechanical Engineering
Design and Construction of a Lighting System to Illuminate a Photobioreactor (103 pp.)
Director of Thesis: Greg Kremer
An internal lighting system for a photobioreactor was designed and constructed
to enhance the growth of green algae. The lighting system is based on a patent for a
solar collector to harness the sun as the primary light source. Fiber optic cables were
fed into the reactor and connected to specially designed light plates that distributed the
light throughout the reactor. The reactor contained Scenedesmus dimorphus, a
microscopic green alga, in a growth media solution. By internally illuminating the
reactor, the energy input cost can be reduced due to the elimination of wasted light.
A 30 liter laboratory scale bubble column reactor was designed and built to test
the objectives laid out. The reactor was designed to be easily assembled and
disassembled, cleaned, and maintained. It was also designed so that the lighting parts
could be exchanged for different configurations if future work necessitates changes.
To design the optimal light plate, the optical simulation software TracePro was
utilized. Information available in literature regarding the absorption and scattering
coefficients of microscopic green algae was analyzed and discussed and the phenomena
of photosynthesis was investigated. A criterion for the selection of the light plate was a
minimum of 20 µEinsteins m-2 s-1 at a distance of 55 cm from the top of the reactor with
a chlorophyll a concentration of 20 mg/L.
Page 4
4
The reactor was inoculated with S. dimorphus and sparged with 1% CO2. The
light intensity was tested using an underwater light sensor in an attempt to validate the
simulation model. Tests also included measurements of pH, chlorophyll a, and turbidity
taken three times daily to study the growth potential of the reactor. The results of the
tests show that the algae grows steadily in the reactor indicating that the proposed
design could be an effective method for growing algae as a petroleum substitute and for
the biologic sequestration of CO2.
Approved:
Greg Kremer
Associate Professor of Mechanical Engineering
Page 5
5
TABLE OF CONTENTS
Page
ABSTRACT ......................................................................................................................................... 3
LIST OF FIGURES ............................................................................................................................... 7
LIST OF TABLES ................................................................................................................................. 9
1 INTRODUCTION ........................................................................................................................... 10
1.1 Energy Demands and Costs Predictions ............................................................................... 10
1.2 Biomass ................................................................................................................................ 12
1.3 Biofuels ................................................................................................................................ 13
1.4 Algae .................................................................................................................................... 14
1.5 Photobioreactors ................................................................................................................. 16
1.6 Project significance .............................................................................................................. 19
1.7 Objectives ............................................................................................................................ 19
2 LITERATURE REVIEW ................................................................................................................... 21
2.1 Carbon Sequestration .......................................................................................................... 21
2.2 Photobioreactors ................................................................................................................. 22
2.3 Bubble Column Reactor ....................................................................................................... 26
2.3.1 Gas Velocity ................................................................................................................... 28
2.3.2 Gas Holdup .................................................................................................................... 28
2.4 Light and dark cycles ............................................................................................................ 29
2.5 Compensation Point............................................................................................................. 34
2.6 Light Transmittance ............................................................................................................. 35
3 DESIGN APPROACH ..................................................................................................................... 38
3.1 Design Parameters and Constraints ..................................................................................... 38
3.1.1 Fiber Optics ................................................................................................................... 38
3.1.2 Gas Sparger ................................................................................................................... 40
3.2 Algae Specimen .................................................................................................................... 40
3.3 Optical Simulations .............................................................................................................. 41
3.3.1 Simulation Software ...................................................................................................... 42
3.3.2 Simulation material properties ..................................................................................... 43
Page 6
6
3.3.3 Light plate design and testing ....................................................................................... 50
3.3.4 Benefits of light plates .................................................................................................. 53
3.5 Illuminator Cooling ............................................................................................................... 64
3.6 Reflective Coating ................................................................................................................ 64
3.7 Test Plan ............................................................................................................................... 67
3.7.1 Analytical Procedures for fluid properties .................................................................... 67
3.7.2 Light Measurements ..................................................................................................... 69
4 RESULTS AND DISCUSSION .......................................................................................................... 73
4.1 Simulation Results ................................................................................................................ 73
4.2 Reactor results ..................................................................................................................... 78
4.3 Scattering due to bubbles .................................................................................................... 90
4.4 Circulation ............................................................................................................................ 91
5 CONCLUSIONS ............................................................................................................................. 94
5.1 Working System ................................................................................................................... 94
5.3 Light Plate Design ................................................................................................................. 95
6 RECOMMENDATION AND FUTURE WORK .................................................................................. 97
6.1 Recommendations ............................................................................................................... 97
6.1.1 Lighting system ............................................................................................................. 97
6.1.2 Reactor components ................................................................................................... 100
6.2 Future Work ....................................................................................................................... 102
WORK CITED ................................................................................................................................. 105
APPENDIX A: Growth material ..................................................................................................... 110
APPENDIX B: Additional plots ...................................................................................................... 111
APPENDIX C: Cost ......................................................................................................................... 113
Page 7
7
LIST OF FIGURES
FIGURE 1. CO2 EMISSIONS OVER TIME FROM COAL FROM 1980 TO 2008 (U.S. ENERGY INFORMATION
ADMINISTRATION, 2010) ............................................................................................................ 11
FIGURE 2. CONVERSION PROCESS FOR BIOMASS TO BIOFUEL (MCKENDRY, 2002) ........................................ 14
FIGURE 3. MICROSCOPIC GREEN ALGAE (WILSON, 2009), LEFT, AND GIANT ALGAE IN NEW ZEALAND (ORIZARSKA,
2007) ...................................................................................................................................... 15
FIGURE 4. ABANDONED HORIZONTAL TUBE FACILITY (MIRON A. S., 1998) .................................................. 17
FIGURE 5. VARIOUS BIOREACTOR SETUPS, THE LOOP REACTOR LABELED HERE IS THE AIRLIFT REFERRED TO IN THIS
WORK (CHISTI, 1983) ................................................................................................................. 18
FIGURE 6. GEOLOGICAL AND BIOLOGICAL SEQUESTRATION METHODS (PLUNGING INTO CARBON SEQUESTRATION
RESEARCH, 2000) ...................................................................................................................... 22
FIGURE 7. NEAR HORIZONTAL TUBE REACTOR (CARVALHO, MEIRELES, & MALCATA, 2006) ........................... 25
FIGURE 8. POSSIBLE FLOWS IN A BUBBLE COLUMN (KANTARCI, BORAK, & ULGEN, 2005) .............................. 27
FIGURE 9. COLUMN DIAMETER AND GAS VELOCITY EFFECTS ON BUBBLE FLOW (KANTARCI, BORAK, & ULGEN,
2005) ...................................................................................................................................... 28
FIGURE 10. COMPARISON OF GROWTH RATES AND CHLOROPHYLL A CONCENTRATIONS WITH RESPECT TO
DIFFERENT LIGHT CYCLE PERIODS (CARR & WHITTON, 1982) ............................................................. 30
FIGURE 11. CELL GROWTH OVER TIME WHEN EXPOSED TO CONSTANT LIGHT FOR APHANOTHECE MICROSCOPICA
(JACOB-LOPES, 2009) ................................................................................................................. 31
FIGURE 12. CELL CONCENTRATIONS WITH DIFFERENT DARK:LIGHT CYCLES (JACOB-LOPES, 2009) .................... 32
FIGURE 13. CO2 ABSORPTION AS COMPARED TO A 24 HOUR PERIOD OF LIGHT (JACOB-LOPES, 2009) .............. 33
FIGURE 14. LIGHT CURVE COMPARING LIGHT INTENSITY TO PHOTOSYNTHESIS RATES (GRAHAM & WILCOX, 2000)
............................................................................................................................................... 35
FIGURE 15. LIGHT BEING REFLECTED AT MEDIA INTERFACE ......................................................................... 35
FIGURE 16. LIGHT REFRACTING (WEISSTEIN, 2007) ................................................................................. 36
FIGURE 17. FIBER OPTIC ILLUMINATOR ................................................................................................... 39
FIGURE 18. FIBER OPTIC POLISHER......................................................................................................... 39
FIGURE 19. SCENEDESMUS DIMORPHUS IN SOLUTION ............................................................................... 41
FIGURE 20. LIGHT ATTENUATION CALCULATIONS PROCESS (BASS, 2010) ..................................................... 45
FIGURE 21. SCATTERING COEFFICIENT CURVES BASED ON DATA FROM TABLE 5 ............................................. 48
FIGURE 22. SHORT (LEFT) AND LONG (RIGHT) VERSIONS OF THE LIGHT PLATE (UNITS IN INCHES) ...................... 51
FIGURE 23. LIGHT PATHS THROUGH THE UNIQUE PLATE DESIGN ................................................................. 53
FIGURE 24. 35° LIGHT PLATE TO BE INSTALLED IN BCR ............................................................................. 56
FIGURE 25. BOTTOM PLATE HOLDER ...................................................................................................... 56
FIGURE 26. TOP PLATE HOLDER ............................................................................................................ 57
FIGURE 27. PARTIALLY CONSTRUCTED COMPONENTS OF BCR .................................................................... 58
FIGURE 28. SLOT MECHANISM FOR LOWER DISK ...................................................................................... 59
FIGURE 29. SLOT MECHANISM FOR UPPER DISK ....................................................................................... 59
FIGURE 30. GAS DELIVERY SYSTEM ........................................................................................................ 60
Page 8
8
FIGURE 31. GAS DELIVERY SYSTEM (LOOKING FROM UNDER THE REACTOR) .................................................. 61
FIGURE 32. INTERNAL LIGHT COMPONENTS PRIOR TO ASSEMBLY WITH COLUMN ........................................... 62
FIGURE 33. LIGHTING COMPONENTS INSTALLED IN COLUMN ...................................................................... 63
FIGURE 34: COOLING SYSTEM FOR ILLUMINATOR ..................................................................................... 64
FIGURE 35. BCR COVERED IN ALUMINUM FOIL ........................................................................................ 65
FIGURE 36. LIGHT PLATES COVERED IN ALUMINUM FOIL EXCEPT OVER THE ANGLED FACES .............................. 66
FIGURE 37. LIGHT SENSOR FIXTURE INCLUDING DEPTH ROD ....................................................................... 69
FIGURE 38. SENSOR ORIENTATION INSIDE REACTOR FOR ACCURATE LIGHT MEASUREMENTS ............................ 70
FIGURE 39. DEPTH ROD WITH MARKINGS SHOWN .................................................................................... 71
FIGURE 40. DISTANCES BETWEEN VARIOUS COMPONENTS OF THE REACTOR TO UNDERSTAND PLACEMENT OF
SENSOR USING DEPTH ROD MARKING. THE SENSOR IN THIS CASE IS AT A DEPTH OF 24 CM. ILLUSTRATION IS
NOT SCALE. ................................................................................................................................ 71
FIGURE 41. RESULTS OF LIGHT SIMULATIONS FOR TWO DIFFERENT CHLOROPHYLL A CONCENTRATIONS
THROUGHOUT THE REACTOR. (A) 35 PLATE, (B) 35 LONG PLATE, (C) 45 PLATE, (D) 45 LONG PLATE ......... 75
FIGURE 42. RESULTS OF LIGHT SIMULATIONS FOR TWO DIFFERENT CHLOROPHYLL A CONCENTRATIONS
THROUGHOUT THE REACTOR. (A) 55 PLATE, (B) 55 LONG PLATE, (C) 60 PLATE, (D) 60 LONG PLATE ......... 76
FIGURE 43. LIGHT INTENSITY DISTRIBUTION (LEFT) AND COLOR MAP (RIGHT) OF THE 35° LONG PLATE AT A DEPTH
OF 425 MM ............................................................................................................................... 77
FIGURE 44. LI-192 UNDERWATER LIGHT SENSOR (LEFT) AND LI-190 LIGHT SENSOR (RIGHT) .......................... 83
FIGURE 45. COMPARISON OF CALIBRATED AND UNCALIBRATED LIGHT SENSOR AT A DISTANCE OF 1.5" FROM LIGHT
SOURCE IN AIR ............................................................................................................................ 84
FIGURE 46. COMPARISON OF CALIBRATED AND UNCALIBRATED LIGHT SENSOR AT A DISTANCE OF 4" FROM LIGHT
SOURCE IN AIR ............................................................................................................................ 84
FIGURE 47. MELTED FIBER OPTIC CABLE AFTER PROLONGED USE WITH THE ILLUMINATOR ............................... 86
FIGURE 48. ABSORPTION COEFFICIENTS FOR WATER CONTAINING PHYTOPLANKTON FOR CHLOROPHYLL A
CONCENTRATION FROM 0.2 AND 18.4 MG M-3 (PRIEUR & SATHYENDRANATH, 1981) ......................... 87
FIGURE 49. CHLOROPHYLL A AND TURBIDITY CHANGES OVER COURSE OF EXPERIMENT. INOCULATION OCCURRED
AT 7 P.M. ON DAY 0 WITH THE FIRST MEASUREMENTS TAKEN ON DAY 1. ............................................. 88
FIGURE 50. DIFFERENCE IN LIGHT PENETRATION WITH RESPECT TO DISTANCE BASED ON BUBBLING. THE POINTS
ARE THE DAY AND TIME OF THE MEASUREMENT, I.E. 1-10 WAS TAKEN AT 10:00 ON THE DAY ONE OF
TESTING. ................................................................................................................................... 91
FIGURE 51. PROGRESSION OF MIXING VIA A DYE TEST ............................................................................... 93
FIGURE 52. LI-193 SPHERICAL LIGHT SENSOR (UNDERWATER PAR MEASUREMENT, 2011) .......................... 98
FIGURE 53. METAL HALIDE FIBER OPTIC ILLUMINATOR (WIEDAMARK, 2011) ............................................... 99
FIGURE 54. LIGHT LOSS OF FROM COUPLER AT INTERFACE WITH LIGHT PLATE .............................................. 101
FIGURE 55. ALTERNATIVE METHOD OF INTERFACING FIBERS WITH LIGHT PLATES BY EMBEDDING FIBERS INTO THE
PLATES .................................................................................................................................... 101
Page 9
9
LIST OF TABLES
TABLE 1. TYPICAL SOURCES OF BIOMASS FOR ENERGY PRODUCTION (STAM, 2009) ....................................... 13
TABLE 2. COMPARISON OF OPEN AND CLOSED ALGAE GROWTH SYSTEMS (UNITED NATIONS, 2009) ................ 23
TABLE 3. COMPARISON OF SEVERAL REACTOR TYPES (CARVALHO, MEIRELES, & MALCATA, 2006) .................. 24
TABLE 4. CHLOROPHYLL A SPECIFIC ABSORPTION COEFFICIENTS WITH RESPECT TO WAVELENGTH (BASS, 2010) .. 46
TABLE 5. SCATTERING COEFFICIENTS (BW) FOR WATER AND SEA WATER (BASS, 2010)................................... 47
TABLE 6. SCATTERING COEFFICIENT WITH RESPECT TO WAVELENGTH FOR PURE SEA WATER ............................. 48
TABLE 7. SCATTERING COEFFICIENTS WITH RESPECT TO WAVELENGTH FOR PURE WATER ................................. 49
TABLE 8. TWO VARIABLE ANALYSIS OF LIGHT INTENSITY USING BEER-LAMBERT'S LAW .................................... 50
TABLE 9. RESULTS FROM 8/16/2011 .................................................................................................... 79
TABLE 10. RESULTS FROM 8/17/2011 .................................................................................................. 79
TABLE 11. RESULTS FROM 8/18/2011 .................................................................................................. 80
TABLE 12. RESULTS FROM 8/19/2011 .................................................................................................. 80
TABLE 13. RESULTS FROM 8/20/2011 .................................................................................................. 81
TABLE 14. RESULTS FROM 8/21/2011 .................................................................................................. 81
TABLE 15. RESULTS FROM 8/22/2011 .................................................................................................. 82
TABLE 16. DIFFERENCE IN MEASUREMENTS WHEN SAMPLES ARE TAKEN FROM THE TOP AND BOTTOM OF THE
REACTOR ................................................................................................................................... 89
Page 10
10
1 INTRODUCTION
1.1 Energy Demands and Costs Predictions
Power generation is a necessary element of today’s society. However, society is
no longer as accepting of the methods used to generate the energy we crave. The older
methods are becoming too expensive and are seen as too great a burden on the
environment. In the United States in 2009 alone energy consumption was 94.87
quadrillion Btu [1]. Although due to economic recession this value is down from 2008
and 2007 when consumption was 99.38 and 101.55 quadrillion Btu, respectively, the EIA
predicts that the energy demand by the United States will steadily increase at an
average rate of .2% a year until 2030 under current trends. While this trend will present
challenges by itself, on the global scale developing countries are quickly taking a larger
share of the global energy supply.
Asia is expected to see the biggest increase, with China leading the way. It is
projected that there will be a 3% yearly increase in energy consumption by Asia alone,
potentially topping 240 quadrillion Btu in 2030 up from just 47.4 quadrillion Btu in 1990.
A great majority of this energy is expected to come from coal. There will be similar
increases in South America where energy demands could rise from 149.9 quadrillion Btu
in 1990 to 400 quadrillion Btu in 2030, again with the biggest increase coming from coal.
This increase in the consumption of coal will lead to an increase in CO2 output.
According to the EIA Carbon Dioxide Emissions report the worldwide output of CO2 was
12,897.9 million tons in 2008, compared to 6,572.5 million tons in 1980. Figure 1 shows
Page 11
11
the emission trends for the last 30 years. This huge spike in emissions is expected to
keep growing due to the industrialization of developing countries where they lack
emissions standards that are in place in the developed world.
This rapidly increasing consumption raises two concerns. The first is that coal,
and all fossil fuels, are nonrenewable. This means that there is a limited supply and that
this supply will become harder to reach and therefore increase in price as the end draws
nearer. The other concern is that CO2 is a greenhouse gas. Greenhouse gases retain
heat that would normally bounce off the surface of the earth and into space. While no
proven effects of this have been observed to be caused by human emissions, it is best to
prevent putting excessive amounts of anything into the air. A method of sequestration
must be developed which is a focus of this work.
Figure 1. CO2 emissions over time from coal from 1980 to 2008 (U.S. Energy Information
Administration, 2010)
0
2000
4000
6000
8000
10000
12000
14000
1975 1985 1995 2005 2015
CO
2 e
mis
sio
ns
( m
illi
on
s o
f m
etr
ic
ton
s)
Year
World
United States
China
Page 12
12
1.2 Biomass
The Biomass Energy Center in the United Kingdom describes biomass as the
“biological material derived from living, or recently living organisms”, generally used
when talking about plant material (Biomass Energy Centre, 2008). Biomass is carbon
based, but does contain organic compounds and elements like oxygen and nitrogen as
well as some other heavy metals. Fossil fuels are concentrated forms of biomass; they
are just not considered “recent” biomass. For the context of this paper biomass will
refer to new matter on a human times scale while fossil fuels will refer to old matter
based on a geological time scale. Biomass is considered a carbon neutral fuel because
when plant material is burned it releases CO2 back into the atmosphere. This CO2 is then
used by plants for growth. This is called the carbon cycle. When biomass is burned and
new plant material is planted and allowed to grow the total carbon in the atmosphere is
relatively unchanged. It is theorized that fossil fuels require millions of years to form;
therefore when fossil fuels are burned, CO2 is released that has been stored for millions
of years and will add to the total carbon in the atmosphere in its current state.
Biomass comes in many forms, from waste material to crops grown specifically
for energy; Table 1 shows many traditional sources of biomass. However this creates a
problem of supply. Quick access to large quantities of plant biomass can be difficult as
only a specific amount is grown or collected. Once this has been used more must be
grown, and for traditional forms of biomass this could take a full season. It is therefore
vital that a new source be found that can be grown quickly and in large amounts. That
Page 13
13
source is believed to be algae; however the process of growing algae still has not been
perfected on an industrial scale.
Table 1. Typical sources of biomass for energy production (Stam, 2009)
1.3 Biofuels
Biomass can be converted to biofuels using natural chemical processes. The
process of anaerobic digestion uses bacteria to break down the biomass in a slurry into
a biogas similar to natural gas. Fermentation is generally done on cereal or sugar plants
and is used to produce ethanol, the process is the same used to make alcoholic drinks.
Composting is similar to anaerobic digestion except that it is typically a dry process.
Figure 2 outlines various processes and results from different conversions of biomass.
Page 14
14
Figure 2. Conversion Process for Biomass to Biofuel (McKendry, 2002)
These methods of producing a renewable source of fuel are very promising
except that they require tremendous amounts of crops that at the present are not
sustainable along with maintaining current food supplies. An example of the
unsustainability of terrestrial biomass, such as sugar or corn, for fuel is that to grow
enough biomass for 1 liter of fuel requires 10,000 L of water (Wijffels & Barbosa, 2010).
However, the use of algae is becoming a popular alternative to traditional biomass
supplies as algae can be grown quickly and with relatively little input. To grow
approximately 1 kg of algal mass requires only .75 L of water for photosynthesis (Wijffels
& Barbosa, 2010).
1.4 Algae
Algae is a broad term that represents a vast grouping of organisms that can
range in size from microscopic to enormous seaweeds. Nearly all algae contain
Chlorophyll a and β-carotene for photosynthesis, and these types of algae are the main
focus of this work. Figure 3 show various types of algae.
Page 15
15
Figure 3. Microscopic green algae (Wilson, 2009), left, and giant algae in New Zealand
(Orizarska, 2007)
Algae produce a large amount of the oxygen in the atmosphere. Algae also serve
as the basis of the earth’s food chain because they are rich in hydrocarbons. Algae were
first looked at as a potential food source that was high in nutrients and quick to grow in
the 1950’s, though they are currently used to make products ranging from cosmetics to
pharmaceuticals to dyes. Some of these products cannot be made any other way.
Research for algae based biofuels intensified in the 1970’s with the energy crisis, along
with other plans for renewable energy (United Nations, 2009).
Along with photosynthesis algae also perform respiration, like other living
creatures, in which oxygen is needed; carbon dioxide is only consumed for
photosynthesis. Respiration occurs constantly, however, it is most noticeable when
photosynthesis is not occurring, such as during dark cycles. During dark cycles, algae
will use some of the growth to sustain itself so there will be a slight loss during these
times. During a full night up to 25% of growth from the previous day can be used,
Page 16
16
depending on temperature, for respiration. Constant lighting has been used to counter
this; however using dark/light ratios to control loss can also be used. Photosynthesis
utilizes less than 50% of the available light spectrum, and it does so very inefficiently.
Algae converts only 11% of light energy to chemical (United Nations, 2009).
1.5 Photobioreactors
Processes that involve microorganisms require reconfiguration of the multiphase
reactor into a bioreactor. A bioreactor must be capable of delivering nutrients to and
removing waste generated by the living organisms. Processes involving bioreactors
include production of penicillin to various enzymes to food products like vinegar, yogurt,
cheese, and beer (Chisti, 1983). These processes do not require light and therefore are
done in the dark. However, new applications of these systems make use of
photosynthetic organisms so the introduction of lighting systems is vital.
There are two types of reactors for growing algae, pond systems and closed
photobioreactors. Both of these systems have distinct advantages and disadvantages.
The major drawbacks to the open systems are the land usage, difficulty in diffusion of
nutrients and CO2, and controlling contamination. Putting nutrients into the slurry can
be a very slow and inefficient process.
Internally lit bioreactors can be made in the form of horizontal tubes or vertical
columns. The horizontal tube systems have been introduced first at a production scale;
however it has proven very difficult to remove excess oxygen from the system.
Nutrients and CO2 are put into the system at one point and waste is removed at the end
Page 17
17
of the tubes. This means that nutrients are usually depleted long before they reach the
end of the tube. The excess oxygen from respiration displaces the CO2 and causes the
algae to starve because it cannot perform photosynthesis. This effect was increased
because mixing is difficult in long horizontal tubes. Tubular systems also require a lot of
land, are difficult and expensive to build and maintain, and have never worked as
theoretically planned, for example the Photobioreactor Ltd facility in Spain was
abandoned because it did not produce as planned. Figure 4 shows a picture of the
tubular structure that was abandoned.
Figure 4. Abandoned horizontal tube facility (Miron A. S., 1998)
The bioreactors that have shown to be the best for reactions requiring light have
been the various vertical column reactors. Typically these kinds of reactors have very
Page 18
18
high gas hold up values and with the bubble column specifically there is a significant
advantage with respect to the light/dark cycles than the horizontal tube reactors (Miron
A. S., 1999). Figure 5 shows some of the variations available with the bubble column.
The bubble column is superior because there is a constant flow of nutrients and waste
material. CO2 is pumped into the bottom of the reactor and allowed to bubble to the
top, as it does so it is consumed by the algae. Also, as O2 is produced it will rise to the
top and will exit the system in an exhaust opening. This bubbling also generates a
mixing current in the slurry which is vital to the algae survival and growth as it
continually brings the algae cells individually into contact with CO2.
Figure 5. Various bioreactor setups, the loop reactor labeled here is the airlift referred
to in this work (Chisti, 1983)
Page 19
19
1.6 Project significance
The Institute for Sustainable Energy and the Environment (ISEE) has been
studying ways to more effectively use non-renewable energy sources and developing
new renewable sources to take the center role of meeting our energy demands. ISEE
has built and tested an airlift style photobioreactor and has need of a bubble column
reactor to do side by side comparisons for algal growth efficiencies. With the bubble
column reactor, future work can continue in the area of carbon sequestration and
petroleum substitutes. This work will also add to the literature information about the
use of lighting in a bubble column reactor, which at the present is greatly lacking.
1.7 Objectives
The main objective of this thesis was to design and construct a lighting system
that will efficiently introduce light into a bioreactor to create conditions under which
algae can grow. To achieve this, a secondary objective was to create a working bubble
column style reactor. This reactor will allow more in depth study of light conditions and
other growth parameters. It is not the objective of this work to achieve a specific
production level for algae growth. Past work was studied on maximizing light diffusion
and penetration through aqueous solutions. A literature review was done to
understand bioreactor design, culturing algae, and the use of internal lighting and how it
can be applied in this situation. The bubble column is comparable in volume to the
airlift system already constructed so that future work can be done on a side by side
comparison as to which design is the most efficient.
Page 20
20
To achieve these objectives, optical simulations were performed to determine
the most efficient light guide design. The light guides were constructed based on the
results and tested in a reactor similar to the simulation model. The light guides were
designed to be capable of producing diffuse light at an intensity value above the
compensation point for green algae. While no specific growth rate will be attempted,
the reactor must show that algae growth is supported.
Page 21
21
2 LITERATURE REVIEW
2.1 Carbon Sequestration
From Figure 1 in the Introduction the trends of CO2 output can be clearly seen to
be increasing. CO2 has been reported to encourage global climate change, and
therefore increasingly stringent restrictions have been placed on burning fossil fuels.
One method to adhere to these restrictions is to sequester, or isolate and store, the CO2
using a means that will not contribute to an increase in atmospheric levels.
Sequestration can take one of two forms; geological or biological. Geological
sequestration involves pumping CO2 into used natural gas pockets or some other empty
area underground. Biological sequestration is using photosynthetic organisms to
remove CO2 from the air. Figure 6 shows both methods. Biological sequestration does
not permanently remove CO2 from the air but instead is intended to “trap” it in a cycle
of continual absorption for photosynthetic use, release by combustion, and then again
absorbed (Kumar, 2010).
However, new research has shown that a more efficient method of biological
sequestration is using algae. Algae sequester CO2 in three different ways: from the air,
from industrial exhaust, or in solid carbonate states (Wang, 2008). Algae naturally take
CO2 from air like all photosynthetic organisms; however air is typically only made up of
0.03-0.06% CO2, which place a limit on algae growth. Industrial flue gas contains
approximately 15% CO2, and can be utilized at little to no additional cost. This source
could be fed directly to the algae or it can be chemically changed to a carbonate and fed
Page 22
22
to algae. A total of 1.8 tons of CO2 are required to grow one ton of algal biomass
(Wijffels & Barbosa, 2010).
Figure 6. Geological and biological sequestration methods (Plunging into Carbon
Sequestration Research, 2000)
2.2 Photobioreactors
The first reactors for growing algae were open ponds systems, both natural and
manmade. Algae from the ocean have been harvested for food when it washes on
shore. These systems relied on solar energy for photosynthesis and for small scale
needs can be practical. However there are some stark disadvantages to this type of
system. In a report from the United Nations open ponds and closed photobioreactors
were compared.
Page 23
23
Table 2. Comparison of open and closed algae growth systems (United Nations, 2009)
Parameter of Issue Open ponds and raceways Photobioreactor (PBR)
Required space High For PBR itself, low
Water loss Very high, may also cause salt
precipitation
Low
CO2-loss High, depending on pond depth Low
Oxygen concentration Usually low enough because of
continuous spontaneous
outgassing
Build-up in closed system requires gas exchange
device (O2 must be removed to prevent inhibition
of photosynthesis and photo-oxidative damage)
Temperature Highly variable, some control
possible by pond depth
Cooling often required (by spraying water on PBR
or immersing tubes in cooling baths.
Shear Usually low (gentle mixing) Usually high (fast and turbulent flows required for
good mixing, pumping through gas exchange
devices
Cleaning No issue Required (wall growth and dirt reduce light
intensity), but causes abrasion, limiting PBR life-
time.
Contamination risk High (limiting the number of
species can be grown)
Low (medium to low)
Biomass quality Variable Reproducible
Biomass concentration Low, between 0.1 and 0.5 g/l High, generally between 0.5 and 8 g/l
Production flexibility Only a few species possible,
difficult to switch
High, switching possible
Process control and
reproducibility
Limited (flow speed, mixing,
temperature only by pond depth)
Possible within certain tolerances
Weather dependence High (light intensity, temperature,
rainfall
Medium (light intensity, cooling required)
Start-up 6-8 weeks 2-4 weeks
Capital costs High ($100,000 per hectare) Very high ($250,000 to $1,000,000 per hectare,
PBR and supporting systems)
Operating costs Low (stirring mechanism, CO2
addition
Higher (CO2 addition, oxygen removal, cooling,
cleaning, maintenance)
Harvesting cost High, species dependent Lower due to high biomass concentration and
better control over species and conditions
Current commercial
applications
5,000 (8 to 10,000) t of algal
biomass per year
Limited to processes for high assed value
compounds or algae used in food and cosmetics
Page 24
24
Photobioreactors (PBRs) differ from open pond systems in that the necessary
conditions for growth, temperature, pH, nutrient supply, etc., can be strictly controlled
and regulated. PBR’s currently rely on solar energy radiating from the exterior of the
reactor for photosynthesis, but this will be required to change as scale up occurs. As
volumes increase external lighting will no longer be capable of supplying energy to all
algae cells as light is absorbed quickly upon entering an algae slurry.
Most PBR’s can be broken down into three categories: tubular, flat plate, and
fermenter reactors (Carvalho, Meireles, & Malcata, 2006). Tubular and flat plate
designs are used most frequently because natural sunlight is used, which has very low
costs associated with it. Fermenters require some kind of internal illumination. Carvalho
provided a comparison of several reactor designs in Table 3. Potential productivity is
also given for various algae species. From looking at this table the fermenter type is
clearly not preferred as it has a poor light harvesting efficiency, difficult scale up, and
very low productivity.
Table 3. Comparison of several reactor types (Carvalho, Meireles, & Malcata, 2006)
Page 25
25
Tubular reactors can be vertical, horizontal, or helical, with most being the first
two. Vertical tubular reactors are usually in the form of airlift reactors or bubble
columns, which is the focus of this work and will be discussed in detail later. Horizontal
tubular reactors pump the algae mixture through a long series of tubes. Some success
has been achieved with this method, however, a major problem of heat generation
arises due to the pumps. Figure 7 shows a generic design for a horizontal tube system.
Figure 7. Near horizontal tube reactor (Carvalho, Meireles, & Malcata, 2006)
Flat plate reactors are narrow panels designed to efficiently use as much
external light as possible. There are several shapes that have been developed and can
be pumped or bubbled. Fermenters have low area-to-volume ratios and therefore low
sunlight harvesting efficiencies. Internally lit fermenters have been developed on a
bench scale, however until an effective method of scale to production is found,
fermenters will not be used for photosynthetic processes.
Page 26
26
2.3 Bubble Column Reactor
Bubble column reactors (BCR) are generally a transparent cylindrical tube with a gas
sparger at the bottom. Gas is pumped into the column through the sparger and is
bubbled up through the system which contains a liquid-solid mixture, called a slurry.
This bubbling provides necessary nutrients to the system as well as mixes the system.
BCR’s have several advantages over other reactor types (Kantarci, Borak, & Ulgen,
2005):
• Excellent heat and mass transfer characteristics
• Low maintenance and operating costs
• Few moving parts
• Active system catalyst addition and removal
• High durability
Bubble columns are not currently being utilized as PBR’s with external lighting
because the column diameter cannot exceed 0.2 m without a severe loss of light
intensity (Miron A. S., 2000). If a suitable scale up could be found to incorporate
internal lighting, the bubble column could be a major alternative to common reactors
today.
There are several flows that are possible with the bubble column; different flows are
useful in different situations. These flows are shown in Figure 8. The homogeneous
flows occur at low gas velocities, typically less than 5 cm/s. A homogeneous flow
consists of uniform bubble sizes that rise at a uniform rate and generate only gentle
mixing. Heterogeneous flows usually occur when gas velocity exceeds 5 cm/s and is
Page 27
27
made up of big and small bubbles. The increase of energy creates more chaotic mixing
and circulation. It has been observed however that the gas-liquid mass transfer is lower
in a heterogeneous flow than homogenous (Kantarci, Borak, & Ulgen, 2005). Turbulent
flow can also introduce additional shear stresses, which may be unacceptable for
organisms, as compared to homogeneous flows. In bubble columns the only sources of
stress are from the bubble movement, which is dependent on gas velocity, and any
lighting structures (Perez, 2006). Flow patterns depend on column diameter and gas
velocity and can be estimated using Figure 9.
The most important parameters to bubble column design are gas holdup, bubble
characteristics, and mass transfer. Gas holdup is the fractional percentage of gas in the
liquid and will be discussed more below. Bubble characteristics are determined by
liquid and gas properties, sparger pore size, superficial gas velocity, and diameter of the
reactor.
Figure 8. Possible flows in a bubble column (Kantarci, Borak, & Ulgen, 2005)
Page 28
28
Figure 9. Column diameter and gas velocity effects on bubble flow (Kantarci, Borak, &
Ulgen, 2005)
2.3.1 Gas Velocity
Superficial gas velocity is the average velocity of gas being bubbled into the
bubble column through a sparger, which is given by Equation 1.
�� � ��
Where Ug = Superficial gas velocity
Q = Volumetric flow rate
A = Cross sectional area
(1)
2.3.2 Gas Holdup
One of the most important design criteria for the bubble column is the gas
holdup value. Gas holdup is defined as the volume fraction of gas that is taken up by
gas bubbles and is given by Equation 2 (Miron A. S., 1999).
Page 29
29
ε � 3.317 × 10�� �P�V���.��
Where � � ���ℎ�� !"
#$%& � '"()*+*)"�,(-*."!/
(2)
The specific power input is the power input divided by the volume of liquid in the
reactor. The power input can be adjusted by changing the rate at which gas is bubbled
into the system. The specific power input can also be defined by equation 3.
P�V� �ρ�gU�
(3)
Where ρL is the density of the liquid phase, g is the gravitational constant, and UG
is the superficial gas velocity as defined earlier. Gas holdup has been shown to increase
as gas velocity and pressure increase and decrease when liquid velocity increases. Gas
holdup is also affected by the column height when the height to diameter ratio of the
column is less than 5:1. When the ratio is larger than this the height effect on holdup is
negligible (Kantarci, Borak, & Ulgen, 2005).
2.4 Light and dark cycles
Photosynthesis is a chemical process that occurs in organisms containing
chlorophyll: terrestrial plants, algae, and some bacteria. These organisms can harness
the energy in a photon to create sugars that they feed on. While light is required for
photosynthetic organisms to survive, a dark cycle is also needed. This dark cycle is
commonly called the Calvin cycle and is when the cell actually uses the photon that it
Page 30
30
captures during the light cycle. During the Calvin cycle the cell goes through the process
of respiration like every creature, in which it consumes oxygen and part of the biomass
to replace chemicals and repair cell structure damaged during photosynthesis.
The Calvin cycle will continue until light is available again so the longer the cell is
in the dark the more mass loss will occur. To counter this many have tried to provide
constant light, however research has shown that constant light does not provide a clear
advantage to intermittent light (Carr & Whitton, 1982; Jacob-Lopes, 2009). Figure 10
shows the relationship between light periods and the average growth rate. For this
experiment the total period is 24 hours so on the right is 24 hours of light and the left is
24 hours of dark. As the process approaches 24 hours of light the effects of the
additional light become less important. This experiment was done with Oscillatoria
agardihii.
Figure 10. Comparison of growth rates and chlorophyll a concentrations with respect to
different light cycle periods (Carr & Whitton, 1982)
Page 31
31
A similar experiment was done by using Aphanothece microscopica with varying
dark/light ratios. Figure 11 shows the cell concentration over time using constant
lighting conditions.
Figure 11. Cell growth over time when exposed to constant light for Aphanothece
microscopica (Jacob-Lopes, 2009)
With constant light the alga reaches a steady state of approximately 5000 mg/L
after about 6 days. This can be compared to other dark/light ratios from the same
experiment shown in Figure 12.
Page 32
32
Figure 12. Cell concentrations with different dark:light cycles (Jacob-Lopes, 2009)
From this it can be seen that when a ratio of 2:22 is used a greater concentration
of algae can be attained than given constant light and a 4:20 ratio still gives an
acceptable concentration curve. Any larger dark aspect and the concentration begins to
diminish quickly with the exception of 12:12, which could be because it is a more natural
and balanced ratio. Both of these experiments give similar information curves for two
different strains of photosynthetic algae therefore these results can give an indication of
how many strains will behave in similar situations.
The other important aspect of growing the algae is not just the biomass
generated but also the CO2 that is removed. Jacob-Lopes also studied how different
Page 33
33
dark/light cycles affected the CO2 absorption. This is for the A. microscopica strain;
however these results can be compared to other photosynthetic organisms because the
underlying process is the same. Figure 13 shows a histogram of the CO2 activity.
Figure 13. CO2 absorption as compared to a 24 hour period of light (Jacob-Lopes, 2009)
This data set is with respect to a 24 hour period of light. As it can be seen the
longer the light is applied the more CO2 will be absorbed. From this it seems that
constant light should be used, however, if the algae is exposed to light constantly it
greatly increases the effects of photoinhibition. This data set has been given as a ratio
Page 34
34
of hours but this same ratio can be applied to seconds. For example 6:18 can be 6 hours
of dark and 18 hours of light and 6 seconds of dark and 18 seconds of light.
2.5 Compensation Point
The compensation point is the light level at which net photosynthesis is equal to
the rate of respiration. “For light level greater than the compensation point, the alga
will make a net gain in photosynthesis over losses due to respiration” (Graham &
Wilcox, 2000, p. 564). The compensation point is shown in Figure 14 at the intersection
point of the gross photosynthetic rate, Pm, and the zero net-O2 exchange line. Figure 14
shows these curves for Chlamydomonas reinhardtii, a green algae. The curves can be
defined by several different equations, two of which are:
3 � 34[1 − exp �:;34�] (4)
3 � 34 tanh(:;34) (5)
Where Pm is the maximum rate of gross photosynthesis, E is the light intensity,
and α is the slope of the initial increase in gross photosynthesis (Jassby & Platt, 1976).
Page 35
35
Figure 14. Light curve comparing light intensity to photosynthesis rates (Graham &
Wilcox, 2000)
2.6 Light Transmittance
When light passes through two different transparent media it will do two things;
part of it will reflect and the rest will refract. Reflection occurs when a beam of light hits
a surface and instead of moving through the surface moves back at the same angle. This
is illustrated in Figure 15.
Figure 15. Light being reflected at media interface
Compensation
point
Zero net-O2
exchange line
Page 36
36
Refraction occurs when the light beam passes through the interface but the
angle it leaves the first media is not the same as its entrance into the second, as shown
in Figure 16. These concepts can be described also by common applications, such as a
mirror for reflection and a lens for refraction. The amount of light that will pass reflect
or refract is determined by the indexes of refraction and the angle of incidence.
Figure 16. Light refracting (Weisstein, 2007)
The indices of refraction are based on how fast light travels through a substance.
The higher the index of refraction the slower light travels through that substance and
when light changes speed it bends back to the normal of the surface. The angle of
refraction can be found using Snell’s Law:
nC sin θC � nG sin θG
Where n1 = index of refraction of exiting media
n2 = index of refraction of entering media
θ1 = angle of incidence
θ2 = angle of refraction
(4)
Page 37
37
If the angle of incidence is changed the amount of light that is reflected or
refracted can be changed, even to the point of total internal reflection, wherein no light
is refracted.
Page 38
38
3 DESIGN APPROACH
3.1 Design Parameters and Constraints
There are several design considerations that must be addressed when designing
a bubble column reactor, the most significant being the availability of light throughout
the reactor. Because algae absorb light and also block light, light only penetrates
approximately 0.2 meters through fluid containing algae. Therefore, if the radius of an
externally lit bioreactor is larger than 0.2 meters there will be a dark zone in the center
as well as dimly lit areas near the center. This is unacceptable for sustainable algae
growth. The reactor can be larger if some kind of internal lighting is utilized, which is
the focus of this work.
The second design consideration is the introduction of CO2 to the system. For
this system in this work, CO2 was sparged into the system in the form of bubbles. To
increase the mass transfer properties based on the gas holdup value of the system,
small bubbles are necessary, because of the longer residence time and greater surface
area.
3.1.1 Fiber Optics
The choice of fiber optics for the light delivery system was based on the long
term goal of using a solar collector to provide natural light. A fiber optic illuminator was
used to simulate the properties of natural light, Figure 17, because the spectrum of the
halogen light bulb is similar to that of solar radiation (Davidson, 1999). The fiber optics
Page 39
39
are 3mm in diameter and are cut to necessary lengths. The ends of the fiber optics are
ground using a glass grinding wheel, Figure 18.
Figure 17. Fiber optic illuminator
Figure 18. Fiber optic polisher
Page 40
40
3.1.2 Gas Sparger
The sparging of gas into the system provides steady mixing of the solution and
introduces CO2. If the gas is sparged too quickly into the system the flow regime may
become heterogeneous, or turbulent, and the algae cells could potentially be harmed or
killed due to increased shear stresses. However, if the gas is sparged too slowly poor
mixing could result as well as suboptimal levels of nutrients (Miron A. S., Bubble-Column
and Airlift Photobioreactors for algal Culture, 2000). Homogeneous flow regime was
designed based on Figure 9 from section 2.3.
The main tube that makes up the column was 8” (0.20 m) in diameter to match
the dimensions of an airlift reactor previously built by the research group. Therefore to
remain in the homogeneous range the superficial gas velocity must be less than 2 in/s (5
cm/s). A sparger from Diffused Gas Technologies was purchased to serve as the
diffuser. This diffuser is a ceramic dome 7 ½” in diameter with a volumetric flow rate
between 0.5 and 2 CFM. This corresponds to gas velocity range of 0.0815 to 0.3262
in/s.
3.2 Algae Specimen
The algae strain used for this work was Scenedesmus dimorphus instead of
Chlorella vulgaris as it was unavailable. S. dimorphus is a fresh water green algae from
the regions of Australia and Saudi Arabia. Figure 19 shows S. dimorphus in solution
taken from the reactor during testing, including its relative size. The change in algae
specimen has a negligible effect on this work. Though the size of C. vulgaris is less than
Page 41
41
half of S. dimorphus, the chlorophyll a is the property being tracked and that is not
based on size of the organism. Also, the data used for the fluid properties was based on
the averages of many different green algae so the data is as valid for S. dimorphus as it is
for C. vulgaris. The scattering affect is also minimal as both species are microscopic and
the majority of scattering are due to macroscopic bubbles as will be shown in section
4.3.
Figure 19. Scenedesmus dimorphus in solution
3.3 Optical Simulations
The most important aspect of the design of the internally lit bubble column is the
lighting system. Software simulations were conducted to determine how light interacts
Page 42
42
with water containing live algae. By using software simulations to understand what is
happening in the reactor, time and money can be saved as many configurations can be
tested without the need to build each. Simulations, however, require a greater
knowledge of the processes and properties than building and testing a light scheme.
Therefore, it is important to determine the most accurate data sets to use for future
work. By comparing light intensities from the real world system to that of the
simulation model, it would be possible to assess the validity of the algae characteristics
used in the simulations.
3.3.1 Simulation Software
To determine the effectiveness of the light plates the optical simulation software
TracePro from Lambda Research was used (Lamdba Research). This program allows for
solid models to be imported from programs such as SolideEdge, and has very basic
modeling functions included in the package. Once a model is entered, the user then
defines the properties of the material and the light source and runs a simulation. The
results of the simulation are based on the scattering and diffraction of light for specific
input geometry, materials and fluid characteristics. The equations of the light models
are based on ray tracing done with the Monte Carlo Method. A Monte Carlo Method
involves random number sequences used to solve otherwise unsolvable sets of
differential equations.
Page 43
43
3.3.2 Simulation material properties
The most important components that need to be defined for the simulation are
the fiber optic cables, the light plates and the algae solution. TracePro includes an
acrylic material in the database so for the fiber optics and the light plates the material
was chosen as such. The algae solution was not included in the database and therefore
was created as part of the modeling work. Optical properties of light are broken into
two categories: apparent optical properties (AOPs) and inherent optical properties
(IOPs). IOPs are dependent on the media whereas AOPs are dependent on the light field
within the media.
Because photosynthesis does not require a specific orientation, AOPs are of
much less significance for this work than IOPs. The most important IOPs are the
absorption coefficient and the volume scattering function (Bass, 2010). Therefore it is
possible to create a simulation version of algae suspension and model photosynthetic
active radiation (PAR) attenuation, Figure 20, if these two properties can be fully
described. For use in TracePro the absorption coefficients, a(λ), and scattering
coefficients, b(λ), are needed in units of mm-1.
Scattering due to bubbles could not be inputted directly into the simulations
because the movement of the bubbles could not be simulated. Only steady state
systems could be worked on. Also, at the time of the simulations the characteristics of
the bubbles were not known because the sparger had not been selected yet. The
scattering due to bubbles does play a role in the dispersion of the light throughout the
Page 44
44
reactor but not in the removal of light from the system. This means that the affect is
not as significant because scattered light is still useful light, as opposed to absorbed light
that is used. The design of the light guide can proceed at this point without bubble
scattering considered because the scattering is consistent regardless of the design.
Page 45
45
Figure 20. Light attenuation calculations process (Bass, 2010)
Page 46
46
The absorption and scattering coefficients are given in the Handbook of Optics
through a review of past work by Andrea Morel that categorizes the coefficients based
on wavelength. Both coefficient sets are given in chlorophyll a specific values. This
allows the values to be used to determine the coefficients for any value of chlorophyll a
concentration. The chlorophyll a specific absorption coefficient was determined by
averaging 14 different phytoplankton from three different studies, Table 4.
Table 4. Chlorophyll a specific absorption coefficients with respect to wavelength (Bass,
2010)
λ (nm) Ac (m2 mg
-1) Λ Ac λ Ac λ Ac
400 .0394 475 .0347 550 .0098 625 .0061
405 .0395 480 .0333 555 .0090 630 .0063
410 .0403 485 .0322 560 .0084 635 .0064
415 .0417 490 .0312 565 .0078 640 .0064
420 .0429 495 .0297 570 .0073 645 .0066
425 .0439 500 .0274 575 .0068 650 .0071
430 .0448 505 .0246 580 .0064 655 .0084
435 .0452 510 .0216 585 .0061 660 .0106
440 .0448 515 .0190 590 .0058 665 .0136
445 .0436 520 .0168 595 .0055 670 .0161
450 .0419 525 .0151 600 .0053 675 .0170
455 .0405 530 .0137 605 .0053 680 .0154
460 .0392 535 .0125 610 .0054 685 .0118
465 .0379 540 .0115 615 .0057 690 .0077
470 .0363 545 .0106 620 .0059 695 .0046
700 .0027
Page 47
47
To determine the scattering coefficient, b, use Equation 5 (Bass, 2010)
H(I) � HJ(I) + 0.30 ∗ �550.NI � ∗ O�.PG (5)
Where C is the chlorophyll a concentration of the algae solution in mg/m3 and bw(λ)
values can be found in Table 18.
Table 5. Scattering coefficients (bw) for water and sea water (Bass, 2010)
λ
(nm)
Pure Water
(10-4
m)
Pure Sea
Water
(10-4
m)
λ
(nm)
Pure Water
(10-4
m)
Pure Sea
Water
(10-4
m)
350 103.5 134.5 475 27.6 35.9
375 76.8 99.8 500 22.2 28.8
400 58.1 75.5 525 17.9 23.3
425 44.7 58.1 550 14.9 19.3
450 34.9 45.4 575 12.5 16.2
475 27.6 35.9 600 10.9 14.1
To find the scattering coefficients for the wavelength range available for the absorption
coefficient the values in Table 2 are plotted, Figure 21 and the equation of the curves
found and used to find bw(λ) for Equation 5.
Page 48
48
Figure 21. Scattering coefficient curves based on data from Table 5
Table 6. Scattering coefficient with respect to wavelength for pure sea water
λ (nm) bw (m-1
) λ bw λ bw λ bw
400 .0781 475 .0398 550 .0203 625 .0103
405 .0747 480 .0370 555 .0194 630 .0099
410 .0714 485 .0364 560 .0185 635 .0094
415 .0683 490 .0348 565 .0177 640 .0090
420 .0653 495 .0332 570 .0169 645 .0086
425 .0624 500 .0318 575 .0162 650 .0082
430 .0597 505 .0304 580 .0155 655 .0079
435 .0570 510 .0290 585 .0148 660 .0075
440 .0545 515 .0278 590 .0141 665 .0072
445 .0521 520 .0265 595 .0135 670 .0069
450 .0498 525 .0254 600 .0129 675 .0066
455 .0476 530 .0243 605 .0123 680 .0063
460 .0455 535 .0232 610 .0118 685 .0060
465 .0435 540 .0222 615 .0113 690 .0057
470 .0416 545 .0212 620 .0108 695 .0055
700 .0053
y = 2.186e-0.009x
R² = 0.9907
y = 2.8598e-0.009x
R² = 0.991
0
0.02
0.04
0.06
0.08
0.1
0.12
0.14
0.16
300 400 500 600
Sca
tte
rin
g C
oe
ffic
ien
t (m
^-1
)
Wavelength (nm)
Pure Water
Pure Sea Water
Page 49
49
Table 7. Scattering coefficients with respect to wavelength for pure water
λ (nm) bw (m-1
) λ bw λ bw λ bw
400 0.0597 475 0.0304 550 0.0155 625 0.0079
405 0.0571 480 0.0290 555 0.0148 630 0.0075
410 0.0546 485 0.0278 560 0.0142 635 0.0072
415 0.0522 490 0.0266 565 0.0135 640 0.0069
420 0.0499 495 0.0254 570 0.129 645 0.0066
425 0.0477 500 0.0243 575 0.0124 650 0.0063
430 0.0456 505 0.0232 580 0.0118 655 0.0060
435 0.0436 510 0.0222 585 0.0113 660 0.0058
440 0.0417 515 0.0212 590 0.0108 665 0.0055
445 0.0398 520 0.0203 595 0.0103 670 0.0053
450 0.0381 525 0.0194 600 0.0099 675 0.0050
455 0.0364 530 0.0185 605 0.0094 680 0.0048
460 0.0347 535 0.0177 610 0.0090 685 0.0046
465 0.0333 540 0.0169 615 0.0086 690 0.0044
470 0.0318 545 0.0162 620 0.0082 695 0.0042
700 0.0040
A simplified version of the calculations done by TracePro is the Beer-Lambert law shown
in Equation 6.
Q � Q�(�R(S)TU Where (I) � �(I) + H(I)
(6)
Comparing a single ray analysis to a multi-ray analysis is not useful, however, as
an understanding of how light attenuates through a medium is important. An example
of the calculations shows this very important phenomena. For example, assume an
initial intensity of 200 µEinsteins at a wavelength of 425 nm. A data table, Table 8, can
be created to show a variety of results based on the two most important criteria,
Page 50
50
distance from the light source and concentration of chlorophyll a. For this example
(I) � 0.00243 and C and l are varied for different situations.
Table 8. Two variable analysis of light intensity using Beer-Lambert's Law
Initial
intensity Distance from light source (mm)
200 10 20 30 40 50
Co
nce
ntr
ati
on
of
Ch
loro
ph
yll a
5 177.14 156.90 138.97 123.09 109.02
10 156.90 123.09 96.57 75.76 59.43
25 109.02 59.43 32.40 17.66 9.63
50 59.43 17.66 5.25 1.56 0.46
75 32.40 5.25 0.85 0.14 0.02
100 17.66 1.56 0.14 0.01 0.00
3.3.3 Light plate design and testing
The light plates used for this work were developed based on a patent for
backlighting LCD televisions (Ciupke & al, 1993). The design concept, Figure 22, takes
advantage of Snell’s law in that the angled faces only allow light interacting with it from
a certain angle to pass through. A key design variable that needed to be determined
through simulation was the angled faces are 35, 45, 55, and 60° based on the angle of
incidence. The common dimension among all the plates is the top edge at 2” and the
Page 51
51
length of the plate various with the change in angle. Also, “long” versions of the plates
were tested to see if this change made a significant difference.
Figure 22. Short (left) and long (right) versions of the light plate (units in inches)
The bioreactor was modeled in Solidedge and imported to TracePro as a .SAT
file. Once in TracePro a light source and an observation disk were added to the model.
The light source was made to mimic the illuminator used in this work. The light consists
of 350,000 rays ranging from 400 to 700 nm. The observation plate is a transparent
object that is included in the model to act as a light sensor. The size of the observation
plate can be changed to encompass the cross section of the column and can be moved
vertically through the column. The light incident upon the surface of the observation
Page 52
52
plate can be selected for analysis as this would represent the light intensity in the
column.
The acrylic light pipes act as a modified light pipe to transport the light but also
to “spread” the light around. The light plate design utilizes Snell’s law by allowing the
light to leave the plate only through the angled faces and then only when the light
interacts with the angled faces at a certain angle. The light rays will “bounce around”
inside the plate until encountering the angled faces as shown in Figure 23.
To be considered a successful design, a light plate must provide at least 20
µEinsteins m-2 s-1 at a distance of 550 mm from the top of the reactor with a chlorophyll
a concentration of 50 mg/L. The light plate that exceeds this specification were
constructed and tested in a real world system.
Page 53
53
Figure 23. Light paths through the unique plate design
3.3.4 Benefits of light plates
There are several benefits that come from using rigid light guides and, more
specifically, the light plate design discussed here as compared to individual fiber optic
cables scattered throughout the reactor. Scattered fibers would provide very
concentrated light intensities at many places throughout the reactor. These intensities,
however, reach levels well above the saturation point for photosynthetic algae for the
space around the end of the fibers. Intensities of this level can cause harm to the algae
through the phenomenon of photohibiton, which from Figure 14, means an end to
photosynthesis caused by damage to the photon receptors which will shut down algae
Page 54
54
growth. Powles reported that light in excess of 300 µEinsteins/(m2s) will shut down
photosynthesis for several hours (Powles, 1984). Hincapie measured light intensities
near the tip of fiber optics to be as high as 600 µEinsteins/(m2s) (Hincapie E. , 2010) and
even with light scattering this intensity is still far above the saturation point. Therefore,
simply installing fibers throughout the reactor is not an efficient method of distributing
light.
The light plate design would take this high light intensity and distribute it over a
great area. The distribution area at the end of 5 fibers is approximately 0.07 in2,
whereas the plates have a distribution area of up to 41 in2. Most of the light passing
through the plate, however, will pass through the angled faces, with a surface area of
3.3 in2 which is nearly 50 times the distributive surface of 5 individual fibers. However,
assuming the light was evenly distributed through the entire plate, the intensity per
square inch would be in the range of 36-73 µEinsteins/in2, depending on the intensity
entering the plate from the 5 fibers. These light levels throughout the reactor will be
much more tolerable to the algae than the burst of light energy delivered by the fibers.
The light plate also provides a path for the light to travel through, which is not
constantly absorbing it. Therefore, more light needed for photosynthesis, blue and red
wavelengths, can be delivered at more locations deeper inside the reactor. The light
distribution for any light guide can be observed with the use of photodiodes.
Photodiodes convert light energy to electrical energy which can be measured. By doing
this all of the light can be accounted for. This will be discussed more in Chapter 6.
Page 55
55
Another benefit is the rigid positioning of the plates that is not as easily achieved
with individual fibers. If not securely fashioned, individual fiber optics would tend to
move based on hydrological forces and this would result in an unpredictable light
distribution and could possibly cause the fibers to pull away from the illuminator. To
secure the fibers to prevent any movement and to ensure that light patterns could be
maintained, there would have to be a substantial structure in place. This structure
would simply take up volume in the reactor as well as introduce complex issues of
shading. The setup utilized for this work only required 1.3L and there were no issues
with shading as all of the support structure were positioned on the outside edge of the
reactor.
3.4 Reactor Design
Based on the simulation results, as discussed below, the layout of the reactor
consists of two layers of four light plates. The light plates follow the 35° design and are
approximately 15” long, as shown in Figure 24. The plates were modeled in Mastercam
X and machined using a CNC milling machine to ensure the accuracy needed for proper
light distribution. The top performing plate in the simulation measured more than 17”
in total length however the plate length had to be reduced to 15” to fit within the
maximum travel on the CNC mill in our machine shop.
Page 56
56
Figure 24. 35° light plate to be installed in BCR
The light plates are held in place using polyester disks, Figures 25 and 26, which
hang on the inside of the column. The bottom disk is designed to allow enough support
to hold the light plate while also allowing for minimum influence on fluid flow.
Figure 25. Bottom plate holder
Page 57
57
Figure 26. Top plate holder
The bottom disk was hung by securing four support rods, Figure 27, to the disk at
the countersink locations. The top circular plate rest on a shoulder machined onto the
support rod. The light plates on the bottom were fed fiber optics via a transport tube
that was dropped down directly on top of them. The support rods were screwed to the
blind flange at the top of the reactor to secure and square them. The transport tubes
pass through the blind flange and are supported by the stoppers located at the top of
the tube.
Page 58
58
Figure 27. Partially constructed components of BCR
The plates are held using a slot and bolt mechanism shown in. This design allows
for easy assembly and removal for cleaning and maintenance purposes. Figure 28
Page 59
59
shows the mechanism for the lower disk and Figure 29 shows the mechanism for the
upper disk. The only difference is the addition of the added step to further secure the
lower plates because the only points of contact for them are the two tabs and a small
slot on the lower disk.
Figure 28. Slot mechanism for lower disk
Figure 29. Slot mechanism for upper disk
Page 60
60
The diffuser chosen for this reactor is the EW-70025-20 model from Diffused Gas
Technologies, Figure 30. This model is a fine bubble air stone that can maintain the
required superficial gas velocity for a homogeneous bubble flow regime. The diffuser
connects through the bottom of the reactor, Figure 31, and provides bubbling to the
entire cross-section of the column.
Figure 30. Gas delivery system
Page 61
61
Figure 31. Gas delivery system (looking from under the reactor)
Page 62
62
Figure 32. Internal light components prior to assembly with column
Page 63
63
Figure 33. Lighting components installed in column
Page 64
64
3.5 Illuminator Cooling
To attempt to minimize the damage to the fiber optics done by the heat from the
illuminator a cooling system was developed utilizing aluminum fins and quartz. Two pieces of
quartz were embedded in the aluminum with an air gap, Figure 34. Also, the upper section of
the illuminator housing was pulled back to allow for forced convection cooling from a fan. The
heat buildup was substantially lower with this added cooling capabilities however, the heat is
still an issue.
Figure 34: Cooling system for illuminator
3.6 Reflective Coating
Through initial testing before inoculation it was observed that light levels
according to the sensor were considerably low compared to expected values from
optical simulations, in a range of 6-8 µEinsteins at the 40 cm mark. It was observed that
light was exiting the light plate through the sides and back of the plates and this is
Page 65
65
believed to be contributing to the low light levels. It was also observed that light was
leaving the column at a much greater rate than was expected from the optical
simulations. In attempts to understand the low light intensity values the reactor was
wrapped in aluminum foil to ensure light remained in the column, Figure 35.
Additionally, the light plates on the top level were also covered in aluminum foil to
direct the light in a way that recreates the simulation conditions, Figure 36.
Figure 35. BCR covered in aluminum foil
Page 66
66
By wrapping the column in aluminum foil, light is much less likely to escape out
of the column and is instead redirected back into the column. The result of this change
is that the average light intensity measured in the reactor at the 40 cm mark is in the
range of 11 to 14 µEisnsteins. The wrapping also acts as a barrier from the growth lights
over the Scenedesmus tanks in the same room so the light measured was light coming
through the light plates.
Figure 36. Light plates covered in aluminum foil except over the angled faces
Page 67
67
The use of aluminum foil on the light plates further increased the light intensity
levels to approximately 20 µmols. The concern with this however is that the aluminum
foil could not be used when the column was inoculated. This situation is discussed
further in Chapter 6.
3.7 Test Plan
To validate the simulation results, the light intensity was measured in a working
reactor along with other vital fluid characteristics. The reactor, 30 L total volume, was
inoculated with six liters of algae solution from an incubator with a chlorophyll a
concentration of 46.4 and turbidity of 16.3 on 8/15/2011, the remainder of the volume
was DI water. 15mL of growth media, the recipe is found in Appendix A, was also added
at this time. The reactor was bubbled with 1% CO2 at a pressure of approximately 1.5
psi with the lights on continuously. Fluid measurements were begun on 8/16/2011 and
continued through 8/22/2011. Testing involved two essential components, the fluid
properties and light intensity.
3.7.1 Analytical Procedures for fluid properties
To determine the characteristics of the algae solution several other tests were
performed on the solution: turbidity, pH, and nitrate and phosphate levels. Samples
were drawn from the top of the reactor via a turkey baster and emptied in a beaker.
Turbidity is the measure of water clarity based on how the material suspended in the
solution affects light passage (Lenntech, 2011). Turbidity can be caused by particulates
such as nitrates and phosphates, bubbles, and dead and living algae cells. A turbidity
Page 68
68
measurement can be used in conjunction with the chlorophyll a measurement to better
understand what the cells are doing as a chlorophyll a measurement does not give
details on the amount of cells. The turbidity of the system was measured using the
Hach 2100P turbidimeter (Hach, 2011)
The limiting reagents in algae growth are usually nitrate and phosphate.
Therefore these compounds must be monitored periodically. To measure these
compounds the Hach DR/890 colorimeter was used. The algae mass was filtered out of
the solution and the remainder was split into two vials. In one vial a compound for
nitrate measurement was added and a compound for phosphate was added to the
other. The vials are then put into the colorimeter for measurement. The colorimeter
uses light to detect mineral concentrations by emitting wavelengths of 420, 520, 560,
and 610 nm through the fluid and detected what light passes through (Hach, 2011). The
nitrate compound contains cadmium which “reduces nitrates present in the sample to
nitrite. The nitrite ion reacts in an acidic medium with sulfanilic acid to form an
intermediate diazonium salt which couples to gentisic acid to form an amber-colored
product (Hach Company, 2005).” and the phosphate compound contains
orthophosphate which “reacts with molybdate in an acid medium to produce a
Phosphomolybdate complex. Ascorbic acid then reduces the complex, giving an intense
molybdenum blue color (Hach Company, 2005).” These colors allow the colorimeter to
measure the concentration of ions in solution.
Page 69
69
The pH was measured using an Accumet pH meter. Because CO2 was bubbled
through the reactor the pH will tend to be more acidic. The algae will tend to keep the
solution basic through their growth processes, however until this point is reached the
pH was stabilized using 1 m NaOH added when the pH falls below 6.8. The remaining
sample after testing was sterilized with bleach and disposed of to prevent
contamination of the reactor.
3.7.2 Light Measurements
To measure the light penetration and intensity in the reactor, a rig was
developed to hold the LI-192 light sensor in the center of the system, Figure 37. The
sensor must be oriented upward, Figure 38, as preliminary tests showed that the sensor
does not measure light coming from behind the sensor.
Figure 37. Light sensor fixture including depth rod
Page 70
70
Figure 38. Sensor orientation inside reactor for accurate light measurements
The light sensor was moved through the reactor via the depth rod. The depth
rod was marked every 2 cm from 24 to 40 cm with respect to the top of the reactor,
Figure 39. When comparing this depth to the depth in the simulation 5 cm must be
added because of the way the simulation was constructed. Therefore, 40 cm
corresponds to 45 cm in simulation. Knowing the intensity at various depths, the light
penetration can be studied as factor of depth and distance from the light source. The
primary light source in the reactor is the surfaces of the light plates. So when the sensor
is lowered it is steadily moving away from the light source. Figure 40 shows distances in
Page 71
71
the upper region of the reactor so to understand where in the reactor the sensor is
when using the depth rod.
Figure 39. Depth rod with markings shown
Figure 40. Distances between various components of the reactor to understand
placement of sensor using depth rod marking. The sensor in this case is at a depth of 24
cm. Illustration is not scale.
Page 72
72
Each set of measurements included the average light intensity taken over ten
seconds with and without bubbling. The only values that could have been compared to
the simulation results are the intensities without bubbling. The simulations were done
without factoring in scattering due to bubbles. The intensities taken can give an
understanding of the impact of bubbling which can then be implemented into future
simulation models.
Page 73
73
4 RESULTS AND DISCUSSION
To determine the effectiveness of the light plate design, simulations were
conducted to study light penetration and attenuation through algae media. Then, the
light plate that performed the best according to the criteria was constructed and
implemented into the system. The system was tested to validate the input parameters
(such as the characteristics of the media) and the output of the simulation.
4.1 Simulation Results
Eight variations of the light plate were tested using the optical simulation
software TracePro. These variations include “short” and “long” versions of plates with
35°, 45°, 55° and 60° angles of incidence as discussed in Chapter 3. The results are
displayed in Figures 40 and 41. While only chlorophyll a concentrations of 50 and 100
mg/m3 are shown on the graph for visibility, concentration of 10 mg/mm3 was also
evaluated. The chlorophyll a concentration can be used as a relationship to the actual
biomass concentration as it describes the production levels of the algae. This is also a
vital measurement as the absorption of light by the chlorophyll a pigment is the factor
with the greatest influence on light penetration and propagation as shown in Tables 4-7.
These results show how well distributed the light from the light plates can be.
For these simulations the minimum light at the 550 mm mark must be greater than 20
µEinsteins/m2s using a chlorophyll a concentration to be a qualifying design. The only
light plate that achieved this was the long version of the 35° plate. This was not an
expected outcome as according to Snell’s law the critical angle for reflection is 60°. This
Page 74
74
is most likely due to the fact the light remained in the plate longer in the 35° than the
60°. Another benefit of the 35° long plate is that it produced the highest light intensity
for 100 mg/mm3 of chlorophyll a at 550 mm. While the difference is relatively small any
additional light distribution is desirable to decrease the amount of energy input needed.
The light distribution for the 35° “long” is also much more conducive to growing
algae as it provides dark areas for the algae. The other plates do not allow sufficient
opportunities to pass through the dark areas. Figures 41-42 shows the light distribution
for the 35° long plate system. This is an important design aspect because many times in
the light distribution there are very bright spots and mostly darkness and this warps the
average. Figure 43 shows a fairly even distribution of light with a few bright spots but
most of the spots are in the 1000-4000 lux range. For these simulations the unit lux was
chosen over W/m2 because it can be more easily compared to µEinsteins on a PAR basis
(Licor, 2010). The other important design criterion is the wavelength of light that
penetrates. Due to the nature of photosynthesis, ranges of blue and red wavelengths
are absorbed and therefore do not penetrate as deep as green light. Figure 43 also
shows the distribution of light based on the wavelength color. The most important
thing to look for in this type of plot is to see if there are blues and reds in that cross
section. The design of the light plate allows for these wavelengths that are absorbed
quickly to be constantly replenished.
Page 75
75
Figure 41. Results of light simulations for two different chlorophyll a concentrations throughout the reactor. (a) 35 plate, (b) 35 long
plate, (c) 45 plate, (d) 45 long plate
Page 76
76
Figure 42. Results of light simulations for two different chlorophyll a concentrations throughout the reactor. (a) 55 plate, (b) 55 long
plate, (c) 60 plate, (d) 60 long plate
Page 77
77
Figure 43. Light intensity distribution (Left) and color map (right) of the 35° long plate at a depth of 425 mm
Page 78
78
4.2 Reactor results
To validate the simulation work, light tests were performed on the functioning
reactor. Light measurements were taken at 10 a.m., 1 p.m., and 4p.m. along with
measurements of pH, turbidity, and chlorophyll a concentration. Because algae
properties can change relatively quickly, measurements taken several times a day and
averaged together is a more accurate way of examining the growth of the algae rather
than taking several measurements at one point in the day.
The results of the measurements are shown in Tables 9-15. The tables show
light intensities measured at steady state (no bubbling) and with homogeneous flow
(with bubbling). It is clear that the scattering due to bubbles does limit the light
penetration, but not at a severe enough rate to disqualify the light plate design. The
diffusive nature of the bubbles may serve to better distribute light to areas that
otherwise may not have received any. The reactor was inoculated on 8/15/2011 and
the first measurements were taken on 8/16/2011. It was decided on 8/17/2011 to take
light measurements with and without bubbling to understand the impact of bubbling.
The intensity was measured in each state without moving the sensor and by just
opening or closing the air regulator to ensure that the sensor location was consistent.
Page 79
79
Table 9. Results from 8/16/2011
Light intensity (µmol m-2 s-1) measured from the top of reactor
Time chlorophyll
a turbidity pH bubbling 40 38 36 34 32 30 28 26 24
µg/L yes/no cm cm cm cm cm cm cm cm cm
10:00 11.6 4.85 5.75 yes 2.61 3.13 3.11 3.46 3.16 3.63 3.72 4.01 4.04
no
1:00 11 5.42 6.8 yes
no 2.41 2.38 2.37 3.9 4.2 4.6 4.92 5.26 5.15
4:00 11.8 4.93 7.041 yes
no 8.78 8.74 9.14 9.45 9.96 9.55 9.78 9.16 9.27
Table 10. Results from 8/17/2011
Light intensity (µmol m-2 s-1) measured from the top of reactor
Time chlorophyll
a turbidity pH bubbling 40 38 36 34 32 30 28 26 24
µg/L yes/no cm cm cm cm cm cm cm cm cm
10:10 13.6 5.13 6.98 yes 5.6 5.75 6.64 7.48 7.59 8.3 8.23 7.51 7.93
no 7.02 7.07 7.81 8.6 8.61 9.14 8.74 8.03 8.13
12:50 13.1 4.97 7.066 yes 5.56 5.79 6.05 6.5 6.82 7.05 6.82 7.02 6.28
no 6.97 7.13 7.29 7.47 7.62 7.67 7.13 6.98 5.9
4:00 14.2 4.77 7.024 yes 6.51 6.67 6.79 7.76 8.38 7.31 8.65 8.75 9.3
no 8.23 8.29 8.25 9.12 9.72 9.93 9.6 10.15 8.97
Page 80
80
Table 11. Results from 8/18/2011
Light intensity (µmol m-2 s-1) measured from the top of reactor
Time chlorophyll
a turbidity pH bubbling 40 38 36 34 32 30 28 26 24
µg/L yes/no cm cm cm cm cm cm cm cm cm
10:30 15.1 4.71 7.004 yes 6.69 6.77 7.01 7.88 8.36 9.57 8.36 9.26 10.17
no 8.09 8.11 8.15 9.11 9.48 10.73 9.2 9.79 10.49
1:00 15.4 4.99 7.06 y 7.48 7.47 7.57 8.64 8.06 8.1 8.99 10.22 10.38
no 9.03 8.96 8.94 10.11 9.21 9.16 9.98 10.86 10.79
4:00 14.6 4.44 6.988 y 9.46 9.53 7.81 8.15 8.56 8.28 8.54 6.88 6.59
n 9.98 10.18 8.62 9.21 9.63 9.52 9.96 8.33 7.92
Table 12. Results from 8/19/2011
Light intensity (µmol m-2 s-1) measured from the top of reactor
Time chlorophyll
a turbidity pH bubbling 40 38 36 34 32 30 28 26 24
µg/L yes/no cm cm cm cm cm cm cm cm cm
10:05 16.6 5.48 6.971 yes 6.03 6.37 6.69 7.2 7.68 8.87 7.8 7.05 7.17
no 7.35 7.6 7.94 8.47 8.69 10.29 8.64 7.65 7.66
1:45 16.6 5.56 7.025 yes 6.74 6.64 6.77 7.71 7.64 7.3 6.94 7.08 6.92
no 7.91 7.92 7.88 8.97 8.6 8.05 7.6 7.79 7.18
4:00 17.3 5.71 6.997 yes 6.55 7.28 8.53 8.86 8.04 7.82 8.35 6.95 6.91
no 7.6 8.44 9.72 9.81 8.86 8.43 8.85 7.39 7.12
Page 81
81
Table 13. Results from 8/20/2011
Light intensity (µmol m-2 s-1) measured from the top of reactor
Time chlorophyll
a turbidity pH bubbling 40 38 36 34 32 30 28 26 24
µg/L yes/no cm cm cm cm cm cm cm cm cm
10:20 18.5 6.39 6.97 yes 5.72 6.12 6.95 7.19 6.98 7.08 7.79 7.7 6.37
no 6.66 7.19 7.99 8.18 7.87 7.83 8.17 8.09 6.45
1:05 20.1 7.27
7.024 yes 6.02 6.35 7.65 7.98 8.49 8.46 8.43 7.68 7.72
no 7.14 7.36 8.56 8.83 8.99 8.82 8.57 7.76 6.98
4:00 19.9 6.3 7.01 yes 5.47 5.92 6.5 6.97 8.21 9.18 7.76 7.87 8.12
no 6.47 6.94 7.47 7.95 9.08 9.92 8.3 8.46 8.55
Table 14. Results from 8/21/2011
Light intensity (µmol m-2 s-1) measured from the top of reactor
Time chlorophyll
a turbidity pH bubbling 40 38 36 34 32 30 28 26 24
µg/L yes/no cm cm cm cm cm cm cm cm cm
10:10 21.3 6.91 7.005 yes 5.44 5.81 6.89 7.15 7.77 7.85 7.41 7.26 7.02
no 6.42 6.83 7.88 8.01 8.47 8.5 7.76 7.39 7.28
1:05 21.3 6.78 7.016 yes 5.57 5.43 5.95 5.99 6.62 7.14 7.5 7.76 7.45
no 6.18 6.34 6.81 6.82 7.32 7.71 7.86 7.59 7.17
4:00 23.8 7.3 7.017 yes 5.24 5.6 5.44 7.17 7.38 7.71 7.24 7.94 7.32
no 6.2 6.5 6.31 8.1 8.15 8.37 7.67 8.27 7.35
Page 82
82
Table 15. Results from 8/22/2011
Light intensity (µmol m-2 s-1) measured from the top of reactor
Time chlorophyll
a turbidity pH bubbling 40 38 36 34 32 30 28 26 24
µg/L yes/no cm cm cm cm cm cm cm cm cm
10:10 22.8 8.54 7.008 yes 4.5 4.89 4.91 5.53 6.42 7.26 7.01 7.19 6.53
no 5.37 5.83 5.67 6.37 7.15 7.94 7.68 7.65 6.89
12:55 22 8.42 7.074 yes 4.59 5.03 5.98 6.23 6.6 7.17 6.23 7.49 6.9
no 5.33 5.93 6.9 7.11 7.4 7.84 6.9 7.92 7.22
4:10 20.8 7.81 7.175
yes 6.75 7.39 7.76 7.44 7.63 7.86 6.24 5.25 4.75
no 6.88 7.64 8.33 8.07 8.42 7.86 7.16 6.17 5.62
Page 83
83
There are several reasons that might explain why the actual test results did not
closely match the simulation results. The first possible source of discrepancy is that the
light sensor has not been calibrated in five years. Licor suggests that their sensors be
calibrated every two years. The sensor was compared to the LI-190 light sensor, Figure
44, in air and measured fairly close values as was seen in Figures 45 and 46; however,
this correlation cannot be applied to the sensor when it is underwater.
Figure 44. LI-192 underwater light sensor (Left) and LI-190 light sensor (Right)
Page 84
84
Figure 45. Comparison of calibrated and uncalibrated light sensor at a distance of 1.5"
from light source in air
Figure 46. Comparison of calibrated and uncalibrated light sensor at a distance of 4"
from light source in air
y = 1.8578e1.3711x
R² = 0.995
y = 1.7594e1.3595x
R² = 0.9932
0
1000
2000
3000
4000
5000
6000
7000
8000
9000
0 1 2 3 4 5 6 7
Lig
ht
inte
nsi
ty (
mic
ro m
ols
)
illuminator marks
uncalibrated
calibrated
y = 0.3718e1.361x
R² = 0.9957
y = 0.3448e1.3675x
R² = 0.9962
0
200
400
600
800
1000
1200
1400
1600
0 1 2 3 4 5 6 7
Lig
ht
inte
nsi
ty (
mic
ro m
ols
)
Marks on illuminator
uncalibrated
calibrated
Page 85
85
The second source of discrepancy is the light patterns in the light plates. In
simulation, most light exited light plates through the angled faces with only negligible
amounts leaving the sides and back. However, in the real world system a great deal of
light leaves the sides and back of the light plates. This is not necessarily a detrimental
aspect of the design; it simply makes a comparison between the simulation model and
the real world system more difficult. Before the reactor was inoculated, the light plates
were wrapped in aluminum foil and the light intensities were measured. With the
aluminum foil acting as a reflector, the light intensity increased by as much as 100%.
Therefore, a reflective coating could bring the light intensity in the target area closer to
the goal intensity.
A third possible source of discrepancy is how the fiber optics connect to the
illuminator. The way the fiber optics pass through the coupler presents the possibility
that some fiber optics are shaded by the others. This means that not all the light is
reaching the reactor whereas in the simulation all of the fibers were given clear line of
sight of the bulb surface. Also, the temperature of the light bulb exceeds the melting
point of the acrylic fiber optics. This results in melting at the tip of the fiber optic, Figure
47. Melted fiber optic tips drastically decrease the amount of light that can be
transmitted through them. This melting occurred after additional heat sinks and cooling
were implemented. These issues are discussed in greater detail in Chapter 6.
Page 86
86
Figure 47. Melted fiber optic cable after prolonged use with the illuminator
A fourth issue could be the data that was used to construct the simulation media
was not appropriate for the strain of algae used or the chlorophyll a concentrations that
were handled in simulation. The absorption coefficient for chlorophyll a could be higher
than was used in the simulations, like the data shown in Figure 48; however I was
unable to run simulations with this new data. If the absorption coefficients are too low
then the penetration will appear higher than it is in the real world. It could also be that
the absorption of light by chlorophyll b played had a higher impact than anticipated.
Both the original and new set of data are just starting points and are not useful in final
analysis and therefore a thorough investigation of the absorption coefficients for a
single strain of algae should be done. This testing will be detailed in Chapter 6. Even
Page 87
87
with the higher absorption coefficient, it is possible to reach the 20 µEinsteins m-2 s-1
criteria established in section 3.5.3 by applying some of the recommendation that will
be presented in Chapter 6.
Figure 48. Absorption coefficients for water containing phytoplankton for chlorophyll a
concentration from 0.2 and 18.4 mg m-3 (Prieur & Sathyendranath, 1981)
Even though the light intensity was lower than expected, the algae grew at a
constant rate. The pH was maintained at approximately 7 with 1% CO2, the
temperature was regulated via the room air handlers, and the nutrient levels were
Wavelenth (nm)
Page 88
88
maintained. This means that the only limiting factor could have been the lighting. The
intensity values were above the compensation point for green algae, which is between 2
and 6 µEinsteins m-2 s-1 (Graham & Wilcox, 2000). Figure 49 shows the changes in
chlorophyll a and turbidity over the time of the experiment.
Figure 49. Chlorophyll a and turbidity changes over course of experiment. Inoculation
occurred at 7 p.m. on Day 0 with the first measurements taken on Day 1.
The chlorophyll a concentration increased due to the changes in the lighting. At
lower light intensities the algae will create more chlorophyll a pigment, this can be seen
in the first three days. Then as the number of algae cells increase the amount of
pigment continues to increase. The turbidity however, is a much more accurate
representation of the cell activitiy than just chlorophyll a. The first three days showed a
y = 1.881x + 9.6857
R² = 0.9729
y = -0.1767x + 5.2656
R² = 0.9547
y = 0.8363x + 2.2727
R² = 0.9604
0
5
10
15
20
25
0 1 2 3 4 5 6 7 8
Ch
loro
ph
yll
a c
on
cen
tra
tio
ns
(µg
/L)
an
d
turb
idit
y m
ea
sure
me
nts
Testing day
average…
Page 89
89
slight decrease in the turbidity of the water which would imply a slight decrease in the
algae concentration. This is expected as the algae adjusted to the new environment.
After the initial adjustment time the algae began to grow at a steady constant rate. The
reator had been run indefinitely it is possible that the chlorophyll a concentration would
have reached and surpassed the 50 µg/L used in the simulations. This proves that the
lighting system does provide lighting in both an intensity and distribution that promotes
algae growth.
The chlorophyll a and turbidity measurements may not be the most accurate
representation of the reactor. A test was performed on different growth tanks also
growing S. dimorphus to determing if taking samples from the top of a reactor resulted
in different values than samples taken from the bottom of a mixed reactor. The results
of these tests are shown in Table 16.
Table 16. Difference in measurements when samples are taken from the top and
bottom of the reactor
Location of sample
Sample #
type of
measurement Bottom Top
1 chlorophyll a 93.3 µg/L 59.1
turbidity 113 38.7
2 chlorophyll a 37 15.1
turbidity 39 20.9
3 chlorophyll a 110.1 52.3
turbidity 250 43.5
4 chlorophyll a 53.5 13.2
turbidity 138 22.5
Page 90
90
While no direct comparison can be made from Table 16, it can be seen that there
is a large difference in values depending on where the sample is taken from even in a
mixed reactor. Knowing that the change is considerable, this makes the growth values
taken from the bubble column reactor much more comparable to other reactors.
4.3 Scattering due to bubbles
An important phenomenon in the study of photobioreactors is the effect of
bubbling on light penetration. To further understand this light intensity was measured
at each depth with and without bubbling. The percent difference of these conditions
changes linearly with respect to depth based on Figure 50. The bubbles do reduce the
penetration depth of the light; however it is important to compare with the steady state
conditions. The same flux that is present in the steady state is within the system with
bubbling, the light is simply directed elsewhere. The main difference in this system from
externally lit systems with regard to bubble scattering is that the light in this reactor is
sourced on the inside, so the light remains in the reactor. With an external lighting
source, light is reflected back out of the reactor and therefore is wasted.
Page 91
91
Figure 50. Difference in light penetration with respect to distance based on bubbling.
The points are the day and time of the measurement, i.e. 1-10 was taken at 10:00 on
the day one of testing.
Because the plot values change linearly during the time that algae was growing
nonlinearly, it can be seen that the bubbles are the major source of light scattering, not
the algae cells. There was a small amount of variability over the week of testing that
could be attributed to scattering due to algae. This short test was not enough to prove
the exact impact of the bubbles, but was important to identify a future area of study.
4.4 Circulation
Circulation was evaluated using a dye test. Dye was injected into the system and
was videotaped while mixing. The dye test was done with all components installed
including the light sensor. However, light was not applied to the system for this test.
The dye is a red fluorescent tablet from Bright Dyes that was dropped from the top of
y = 0.6647x - 10.629
R² = 0.9666
-15
-10
-5
0
5
10
15
20
25
30
23 25 27 29 31 33 35 37 39 41
Pe
rce
nt
dif
fere
nce
Distance from top of reactor (cm)
17-10 17-1 17-4 18-10 18-1
19-10 19-1 19-4 20-10 20-1
20-4 21-10 21-1 21-4 22-10
22-1 22-4 average
Page 92
92
the reactor and allowed to sink (Bright Dyes, 2011). The video shows that mixing
beginning along the path of the dye tablet but then moves to the center of the reactor
and radiates outward. The entire reactor is mixed in approximately 46 seconds as
shown in Figure 51.
The circulation of the fluid was important for two main reasons. The first was
the exchange of matter from between the algae and the water. If the cells were not
moved they would not have access to the nutrients in the system as well as the CO2.
The second reason was to move the algae in and out of lighted areas. The reactor was
run with the illuminator on constantly. Photosynthesis however requires time when the
light intensity is below the compensation point that was discussed in section 2.5. The
moving fluid carried the cells in and out of contact with the light. During the test, eddy
currents could be seen that created a twisting motion through the solution. These eddy
currents appeared to be random in their appearance and direction.
Page 93
93
Pre-test condition
Dye tablet added to
reactor
Time: 0 seconds
Bubble reach the
top of reactor
Time: 10 seconds
Bottom half of
reactor is mixed
Time: 22 seconds
Reactor is fully mixed
Time: 46 seconds
Figure 51. Progression of mixing via a dye test
Page 94
94
5 CONCLUSIONS
The objective of this work was to develop a functioning bubble column
bioreactor with an internal lighting system in which algae could grow. To achieve these
objectives, optical simulations were conducted and a reactor was built based on the
results of the simulations. The design specifications for the simulations was a minimum
of 20 μEinsteins m-2 s-1 at a distance of 55 cm with a chlorophyll a concentration of 50
mg/L to create an environment in which algae can grow.
5.1 Working System
The proposed flow regime for the reactor was homogeneous based on the
literature review and Figure 9. To achieve this, an air pressure of 1.5 psi was used with a
mixture of shop air and 1% CO2. The result was a homogeneous flow with small bubble
size. The homogeneous flow was also evident from the dye test discussed in section
4.4. The small bubbles increased the gas holdup of the system and the mixing
capabilities with clearly evident eddy currents that appeared random in occurrence,
position, and direction of flow. The circulation resulting from this fluid flow is consistent
with reactors found in literature and is suitable for sustaining algae growth. There was
also only negligible algae adherence to the component surfaces.
The algae growth, given in terms of chlorophyll a and more importantly turbidity,
Figure 49, shows that when the testing ended the algae was still in a growth phase. The
numbers appear low when compared to the airlift reactor or to the other incubators;
however, as was discussed in section 4.2 this could be due to the location of the sample.
Page 95
95
5.3 Light Plate Design
The specification for the light plate design during simulation was that an array of
4 plates must provide at least 20 μEinsteins m-2 s-1 at a distance of 55 cm from the top of
the reactor filled with a solution with a chlorophyll a concentration of 50mg/L. The only
light plate arrangement that met this criterion was the 35° “long”, shown to give an
intensity of approximately 25 μEinsteins m-2 s-1 in Figure 41 in section 4.1. The light
intensity curves followed an expected pattern that decreased quickly and leveled off the
deeper the sensor was located. In the simulation the principal method of light
dispersion from the light plate was through the angled faces toward the middle of the
reactor, Figure 23 in section 3.3.3. This was the expected light path based purely on
Snell’s law and a simplistic view of the interaction between light, acrylic, and the growth
media.
The results from the real world reactor testing were given in Tables 9-15 in
section 4.2. These values were lower than expected from the simulation results with
the highest intensities being around 10 µEinsteins m-2 s-1 at a chlorophyll a
concentration of 15.1 mg/L. I was unable to test a chlorophyll a concentration of 50
mg/L as the solution never got to this level as the highest chlorophyll a concentration
was attained 23.8 mg/L. I was therefore, unable to meet the objective of replicating the
simulation results in the real world system. The light sensor used in this experiment cast
some doubt on the measurements taken due to the type of sensor and the number of
years since it was calibrated. Also, the simulation results were very rigid, as no light left
Page 96
96
the light plates through the sides; all of it exited through the angled faces. In the real
world system, light did leave through the sides and the back. Therefore, there was not
necessarily less light in the system, it simply did not match the simulation model as the
light was more diffuse and filled the reactor rather than just lighting the middle. There
are other possible reasons for this behavior and these were addressed in section 4.2.
Nevertheless, the light intensity measured in the reactor never fell below the
compensation point for green algae as discussed in section 2.5
As was described in Chapter 4, light was controlled to be the limiting reagent.
However, because the algae grew under the presented conditions, the lighting system
performed successfully regardless of the inability to replicate the simulation results.
The reactor met all of the requirements for sustaining algae growth by means of proper
mixing and CO2 availability. Also, knowledge and experience gained from this work
regarding internal lighting systems will add important information to the scarce amount
of data available in this field.
This project achieved all of the stated objectives and ISEE now has a working
internally lit bubble column reactor to compliment the internally lit airlift reactor for
future studies.
Page 97
97
6 RECOMMENDATION AND FUTURE WORK
6.1 Recommendations
The following are modifications or additions to the reactor that should be
considered for any future work. Following these recommendations the reactor could
function more in line with the simulation results. The recommendations include
changes to the lighting system and component materials that have been identified
during this work.
6.1.1 Lighting system
The first and most important recommendation for future work on this project
would be the purchase of an LI-193 Spherical underwater sensor from Licor, Figure 52.
This is the only sensor of its kind available on the market except for very large light
sensors that are used in ocean research. The LI-193 allows for a better collection of light
in algae solutions as it can sense light from all directions (Underwater PAR
Measurement, 2011). The LI-192 can detect light incident from the hemisphere above
the sensor surface, however this is a severe limitation in setup as it has to be located in
the exact same position and orientation each time. Any disturbance will change the
value read. This limitation has also been noted by a group of German optical scientists
when they compared various light sensors (Hartig and Heinrich, 1999). By using the LI-
193, the orientation of the sensor is negligible especially when exposed to diffuse light.
The LI-193 could be implemented in the same way in the reactor that the LI-192, with
Page 98
98
the bulb up as this has been shown in this work to be a stable method of holding the
sensor.
Figure 52. LI-193 Spherical Light Sensor (Underwater PAR Measurement, 2011)
Also, the illuminator needs to be modified or a better light source needs to be
found The light opening on the illuminator is too small to allow all the fiber optics the
appropriate lighting area. The heat from the bulb also degrades the fiber optics and
prohibits light transmission. An improvement that still uses the full visible range of the
light spectrum, to simulate natural sun light, is a metal halide. The metal halide has
several very important advantages to the halogen. A 150W halogen has an efficiency of
18-20 lumens/Watt, whereas a 150W metal halide bulb has an efficiency of 65-115
lumens/Watt. The metal halide also illuminates with a color temperature more light the
sun (Venture Lighting International, 2011) which Hincapie discussed as an important
parameter on algae growth (Hincapie E. , 2010). A halide bulb also operates cooler than
a halogen bulb (Edmund Optics, 2011). The cost of implementing such a system would
Page 99
99
be similar to the halogen illuminator presently in service; however, the bulbs are more
expensive. Figure 53 shows a fiber optic halide illuminator that is recommended for
future work with this reactor. It is a 150W illuminator that comes with an on/off
remote.
Figure 53. Metal halide fiber optic illuminator (Wiedamark, 2011)
A more detailed analysis of the potential of the light plates needs to be
performed as well, especially comparing to individual fibers. This needs to be done to
account for all the light leaving the plates and in what distribution. A method of
determining the light distribution for the optics and light plate comparison would be to
construct a box filled with photodiodes on the inside walls. Photodiodes convert light
energy to electric energy in the form of current or voltage which could be read with a
simply multimeter, or set up the entire array to be measured using a visual
representation on a computer screen. This method would be much more efficient than
using the simulation software TracePro again.
Page 100
100
6.1.2 Reactor components
Another modification to this system should also be to find a method of applying
a reflective surface to the light plates. As stated in Chapter 3, light was leaving surfaces
unexpectedly when compared to the optical simulations. A reflective coating was not
believed to be necessary after initial testing outside of the reactor, however, when the
aluminum foil covering was applied to the plates the amount of light incident on the
sensor increased nearly 125%. A possible coating would be a dielectric mirror coating
from Evaporated Coatings Inc. It is the author’s recommendation that a reflective
surface be applied to the surface closest to the reactor wall and not to the side of the
plate. By coating the back surface of the plate, the system will mimic the simulation
results to a greater degree. From observation of the functioning system, it is believed
that leaving the sides exposed is beneficial to the overall function of the system by
providing more and larger light zones.
Another change to the components should be a solid color fiber coupler to
attach to the light plates. The clear plastic was used for the prototype so that the
interface of the fibers and the light plates could be observed. However, this allowed a
substantial amount of light to escape before entering the light plates, Figure 54. A lot of
the light at the interface of the fibers and plate did not enter the plate but instead
reflected through the coupler. By using a solid white plastic such as a polyester or
acrylic instead of the clear acrylic, nearly all the light can be focused into the light plates.
Page 101
101
Figure 54. Light loss of from coupler at interface with light plate
Another solution to this problem could be to make the light plates slightly wider
and drill holes into the top of the plates and insert the fiber optics directly into the
plate, Figure 55. With this setup more light will enter the light plates since the tips are
embedded in the plate itself. The light that would have reflected out will have a high
chance of being refracted into the plate, either through the base of the hole or the walls
of the hole, and working down through the plate. With the modifications given here it is
believed that the reactor will be more capable of achieving the objectives laid out in
section 1.7. Again, a reflective coating could also be applied near the top of the plate on
all sides to contain more light in the plate.
Figure 55. Alternative method of interfacing fibers with light plates by embedding fibers
into the plates
Fiber optics
Light plate
Page 102
102
The first change should be the implementation of the white coupler. Drilling
holes into acrylic can difficult and the inside of the holes must be clear of any marks.
The glazing process used on the exterior of the light plates will probably not work on the
holes as the heat will not reach the bottom of the holes before damage is done to the
top surface of the plate. Therefore, Figure 55 should only be attempted if the solid
white coupler does not provide enough light control.
These changes to the fiber coupler component will also affect the light
distribution inside the light plate. The method used for this work may have produced
bright and dim spots in the light plates corresponding to fiber location instead of the
even distribution that was expected. By making these modifications to the coupler, the
light will be distributed much more evenly and should eliminate, or significantly
decrease the effect or, bright spots.
6.2 Future Work
A long term testing phase is recommended now that the system has proven it
can function properly. This work was involved in the design and construction only and
therefore was not concerned with long term growth. But for this system to be a
potential bioreactor for algae, the long term capabilities must be known. Growth rates
should be studied using varying CO2 content and light cycles over periods of weeks and
months. For long term testing, the reactor also needs to be modified so that the reactor
can drain from the bottom. The prototype was developed without continual growth
Page 103
103
considered and for long term growth studies the reactor will need to be harvested
during testing based on turbidity measurements.
Developing a controller and sensor to change the light and CO2 levels in real time
would be very helpful to prevent over or under saturation. This could be done by
attaching CO2, pH, and light sensors to MATLAB and use SimuLink and a feedback loop
attached to an injection system and the lighting system. Using an active response
system would provide optimal growth conditions at all times which would be important
in a full scale up.
A method of determining the benefit of the light plate design, or any light guide,
is to use photodiodes to measure light intensities and distributions. Photodiodes (PDs)
are the main components in commercial light sensors and are available through most
electric vendors. A box made of clear acrylic should be constructed with dimensions of
6”x3”x18” with PDs positioned around the outside of the box. The box will have to be
filled with water or even an alga solution as the light patterns are different for wet and
dry conditions. If possible, bubbling should also be done but this could be difficult in a
small volume to create the same bubble characteristics that would occur in a larger
reactor. The light guides can be inserted into the box and light passed through it. The
PDs can be run to a computer and a visual representation of the light intensity in the
box can be shown on a screen. The more PDs used, the greater accuracy the images will
have. Therefore, the entire surface of the box should be covered with PDs. This will
show the benefits of the light plates over single fibers in terms of optimal light usage.
Page 104
104
Long term testing should also include better understanding of the absorption of
chlorophyll a. Sample should be taken with a known chlorophyll a concentration and
the light attenuation should be measured and this data set should be used in
simulations. To do this a tool such as the spectronic 20D from Thermo Scientific
(Thermo Scientific) is needed. This device measures absorptance of a sample and thus,
knowing the chlorophyll a concentration, a complete set of absorption coefficients can
be determined. This data set would be much more accurate for the system in use and
could be used for improvements to the light system design.
Page 105
105
WORK CITED
Plunging into Carbon Sequestration Research. (2000). Retrieved September 8, 2010, from Oak
Ridge National Laboratory:
http://www.ornl.gov/info/ornlreview/v33_2_00/research.htm
LI-192 support page. (2009). Retrieved from LI-COR:
http://envsupport.licor.com/index.jsp?menu=Light_Meters_%26_Sensors&spec=LI-192
Underwater PAR Measurement. (2011). Retrieved 2011, from LICOR:
http://www.licor.com/env/products/light/underwater.jsp
Anderson, G., & Kommareddy, A. (2002). Photobioreactor Design. Society for Engineering in
Agricultural, Food, and Biological Systems. Saskatoon.
Barsanti, L., & Gualtieri, P. (2006). Algae: Anatomy, Biochemisty and BioTechnology. Boca Raton:
CRC Press.
Bass, M. (2010). Handbook of Optics: Optical Perperties of Materials, Nonlinear Optics, Quantum
Optics (3 ed., Vol. 4). The McGraw-Hill.
Baxter, L. (2004). Biomass-coal co-combstion: opportunity for affordable renewable energy.
Fuel, 1295-302.
Bayless, D. (2007). Patent No. 746,389. United States of America.
Biomass Energy Centre. (2008). What is Biomass. Retrieved November 17, 2010, from Biomass
Energy Centre:
http://www.biomassenergycentre.org.uk/portal/page?_pageid=76,15049&_dad=portal
&_schema=PORTAL
Bontanicare. (n.d.). Cal-Mag. Retrieved 2011, from
http://www.americanagritech.com/supplements/cal-mag
Bontanicare. (n.d.). Pro Grow. Retrieved 2011, from
http://www.americanagritech.com/fertilizers/pure-blend-pro/pure-blend-pro-grow
Bright Dyes. (2011). Retrieved 2011, from http://www.brightdyes.com/
Carr, N. G., & Whitton, B. (1982). the Biology of Cyanobacteria. Los Angeles: Blackwell Scientific
Publications.
Carter, B. (n.d.). Coal Availability Studies: a new look at resource estimates. Reston: U.S.
Geological Survey.
Page 106
106
Carvalho, A. P., Meireles, L. A., & Malcata, F. Z. (2006). Microalgal Reactors: A Review of
Enclosed System Designs and Performances. Biotechnology Progress, 1491-1506.
Chisti, M. (1983). Airlift Bioreactors. New York: Elsevier Science Publishing Co.
Ciupke, W., & al, e. (1993). Patent No. 5485354. United States.
Conroy, P. J. (1997). Fluidized Bed Reactor. Retrieved August 6, 2010, from Rensselaer:
http://www.rpi.edu/dept/chem-eng/Biotech-Environ/IMMOB/fluidized-bed.htm
Davidson, M. W. (1999). Tungsten-Halogen and Incandescent Lamps. Retrieved 2011, from Zeiss:
http://zeiss-campus.magnet.fsu.edu/articles/lightsources/tungstenhalogen.html
Demirbas, A. (2005). Potential applications of renweable energy souces, biomass combustion
problems in boiler power systems and combustion related envirnomental issues.
Progress in Energy and Combustion Science, 171-92.
Dudukovic, M. (1999). Multiphase Reactors: Revisited. Chemical Engineering Science, 1975-95.
Edmund Optics. (2011). Dolan-Jenner MH-100 Metal Halide Fiber Optic Illuminator Model.
Retrieved 2011, from
http://www.edmundoptics.com/products/displayproduct.cfm?productid=2365
Graham, L. E., & Wilcox, L. W. (2000). Algae. Upper Saddle River: Prentice Hall.
Graham, L., Graham, J., & Wilcox, L. (2009). Algae. San Fransico: Pearson Benjamin Cummings.
Hach. (2011). 2100P Is Portable Turbidimeter. Retrieved 2011, from
http://www.hach.com/2100p-is-portable-turbidimeter/product?id=7640450955
Hach. (2011). DR/890 Portable Colorimeter. Retrieved 2011, from http://www.hach.com/dr-890-
portable-colorimeter/product?id=7640439041&callback=qs
Hach. (2011). DR/890 Portable Colorimeter. Retrieved 2011, from Hach Company:
http://www.hach.com/dr-890-portable-colorimeter/product-details?id=7640439041
Hach Company. (2005). DR/890 Datalogging Colorimeter Handbook.
Haq, Z. (n.d.). Biomass for Electricity Generation. Energy Information Administration.
Hincapie, E. (2010). Design, Construction and Validation of an Internally-Lit Airlift
Photobioreactor. Athens: Ohio University.
Jacob-Lopes, E. (2009). Effect of light cycles (night/day) on C)2 fixation and biomass production
by microalgae in photobioreactors. Chemical Engineering and Processing, 306-10.
Page 107
107
Jassby, A. D., & Platt, T. (1976). Mathematical formulation of the relationship between
photosynthsis and the light for phytoplankton. Limnology and Oceanography, 540-47.
joshi, J. (2001). Computational Flow Modeling and Design of Bubble Column Reactors. Chemical
Engineering Science, 5893-933.
Kantarci, N., Borak, F., & Ulgen, K. (2005). Review: Bubble Column Reactors. Process
Biochemistry, 2263-83.
Kommareddy, A. (2003). Design and construction of light guides for efficient transer of light in to
a dense algal culture in a Photo-Bio Reactor (PBR). Society for engineering in
agricultural, food, and biological systems. Las Vegas: ASAE.
Kommareddy, A., & Anderson, G. (2003). Mixing and Eddie Currents in a Modified Bubble
Column Reactor. CSAE/ASAE. Fargo: The Society for Engineering in Agricultural, Food,
and Biological Systems.
Kommareddy, A., & Anderson, G. (2004). Study of light requirements of a photobioreactor.
ASAE/CSAE. Winnipeg.
Krishan, R., & Sie, S. (n.d.). Design and Scale-up of the Fischer-Trpsch Bubble Column Slurry
Reactor. Fuel Processing Technology, 73-105.
Krishna, R., & Jurg, E. (1996). Gas Goldup in Bubble Column Reactors Operating in the Churn-
Turbulent Flow Regime for Bubble Column Reactors. AlChE Journal.
Kumar, A. (2010). Enhanced CO2 Fixation and biofuel production via mictoalgae: recent
developments and future directions. Cell Press, 371-80.
Lamdba Research. (n.d.). TracePro suite of optical and illumination design software. Retrieved
from Lamda research corporation:
http://www.lambdares.com/software_products/tracepro/
Lenntech. (2011). Turbidity. Retrieved 2011, from http://www.lenntech.com/turbidity.htm
McKendry, P. (2002). Energy Production from Biomass (part 1): Overview of Biomass. Biosource
Technology, 37-46.
McKendry, P. (2002). Energy Production from Biomass (part 2): Conversion Technology.
Bioscience Technology, 47-54.
McKendry, P. (2002). Energy Production from Biomass (part 3): Gasification Technologies.
Bioresource Technology, 55-63.
Page 108
108
Miron, A. S. (1998). Algae-Based Biofules: A Review of Challenges and Opportunities for
Developing Countrie. Journal of Biotechnology, 249-270.
Miron, A. S. (1999). Comparative Evaluation of Compact Photobioreactors for Large-Scale
Monoculture of Microalgae. Journal of Biotechnology, 249-70.
Miron, A. S. (2000). Bubble-Column and Airlift Photobioreactors for algal Culture. AlChE Journal,
1872-87.
Molina, E. (2001). Tubualr photobioreactor design for algal cultures. Journal of Biotechnology,
113-32.
Navin Biotech. (2010, April 4). Retrieved August 5, 2010, from Himfr: http://www.himfr.com/d-
p1116600355018790225-Trickle_Bed_Reactor/
Navin Biotech. (2010, September 3). Trickle Bed Reactor. Retrieved September 3, 2010, from
Qualitydress:
http://www.google.com/imgres?imgurl=http://img.alibaba.com/photo/11796946/Trickl
e_Bed_Reactor.jpg&imgrefurl=http://www.qualitydress.com/d-
p1116600356307176900-
trickle_bed_reactor/&usg=__g0kXbs4B8XJmZ7QvF3ZQK9ZuxtQ=&h=600&w=395&sz=96
&hl=en&start=0&zoom=1&t
Nussbaumer, T. (2003). Combustion and Co-combustion of Biomass: Fundamentals,
Technologies, and Primary Measures for. Energy and Fuels, 1510-21.
Orizarska. (2007, October 8). Retrieved 2010, from TravelPod:
http://www.travelpod.com/travel-blog-
entries/orizarska/worldtraveler/1202456820/tpod.html
Perez, J. A. (2006). Shear rate in stirred tank and bubble column bioreactors. Chemical
Engineering Journal, 1-5.
Powles, S. B. (1984). Photoinhibition of photosynthesis induced by visible light. Annual review of
plant biology, 15-37.
Prieur, L., & Sathyendranath, S. (1981). An optical classification of coastal and oceanic waters
based on the specific spectral absorption curves fo phytoplankton pigments, dissolved
organic matter, and other particulate materials. American Society of Limnology and
Oceanography, 671-89.
Rhen, C. (2006). Effect of raw material composition in woody biomass pellets on combustion
characteristics. Boimass and Bioenergy, 66-72.
Page 109
109
Richmond, A. (1993). A New Tubular Reactor for Mass Production of Microalgae Outdoors.
Journal of Applied Phycology, 327-32.
Shah, Y., Kelkar, B., & Godbole, S. (1982). Design Parameters Estimations for Bubble Column
Reactors. AlChE Journal, 353-79.
Stam, A. (2009). Review of models and tools for slagging and fouliong prediction for biomass co-
combustion. Arnhem: University of Twente.
The University of Texas at Austin. (n.d.). Media Details. Retrieved august 26, 2010, from UTEX
The Culture Collection of Algae:
http://www.sbs.utexas.edu/utex/mediaDetail.aspx?mediaID=26
Thermo Scientific. Apectronic 20D+ Spectrophotometer. Retrieved 2011, from Thermo Scientific:
http://www.thermoscientific.com/ecomm/servlet/productsdetail?productId=11953277
&groupType=PRODUCT&searchType=0&&storeId=11152
U.S. Energy Information Administration. (2010). CO2 emissions from the consumption of coal
(million metric tons). Department of Energy.
United Nations. (2009). Algae-Based Biofules: A Review of Challenges and Opportunities for
Developing Countries. Rome.
Venture Lighting International. (2011). The Metal Halide Advantage . Retrieved 2011, from
Venture Lighting: http://www.venturelighting.com/techcenter/metal-halide-
techintro.html
Wang, B. (2008). CO2 bio-mitigation using microalgae. Applied Microbiology and Biotechnology,
707-18.
Weisstein. (2007). Snell's Law. Retrieved November 16, 2010, from Wolframresearch:
http://scienceworld.wolfram.com/physics/SnellsLaw.html
Wiedamark. (2011). Retrieved from
http://www.wiedamark.com/150wattmetalhalideilluminatorwith8colorwheel.aspx
Wijffels, R. H., & Barbosa, M. J. (2010, August 13). An Outlook on Microalgal Biofuels. Science,
pp. 796-99.
Wilson, P. (2009, March 3). Information about spirulina, chleorella and blue-green algae.
Retrieved 2010, from Momentum98:
http://www.momentum98.com/spirulinahistory.html
Page 110
110
APPENDIX A: Growth material
The growth media for the experiment was a 6:1 ratio of Botanicare Pro Grow and Botanicare
Cal-Mag.
Page 111
111
APPENDIX B: Additional plots
5.5
5.7
5.9
6.1
6.3
6.5
6.7
6.9
7.1
7.3
0 5 10 15 20 25
pH
Test number
Page 112
112
5
6
7
8
9
10
11
12
22 24 26 28 30 32 34 36 38 40 42
Lig
ht
inte
nsi
ty (
µE
inst
ein
s m
-2s-1
)
distance from the top of reator(cm)
13.6
13.1
14.2
15.1
15.4
14.6
16.6
16.6
17.3
18.5
20.1
19.9
21.3
21.3
23.8
22.8
22
20.8
Page 113
113
APPENDIX C: Cost
Total cost of project
Item quantity units
unit
cost total cost
$ $
8" clear pvc 3.5 feet $44.50 $155.75
hardware
3/4"-10 bolts 8 unit $9.85 $78.80
3/4"-10 nuts 8 unit $0.46 $3.68
miscellaneous
hardware
$20.00
plastic
components
acrylic
12"x12"x1/4" 2 unit $14.60 $29.20
acrylic
24"x24"x1/4" 2 unit $46.20 $92.40
polyester
12"x12"1/4" 2 unit $8.03 $16.06
polyester tube
ID 3/8" OD
1/2" 6 feet $0.90 $5.40
polyester rod
1/2" 7 feet $1.95 $13.65
Flanges
blind flange 2 unit $136.00 $272.00
open flange 2 unit $123.00 $246.00
gasket 1 unit $11.00 $11.00
Sparger 1 unit $50.00 $50.00
man hours
CNC machinist 6 hours $20.00 $120.00
general shop 4 hours $10.00 $40.00
total $1,153.94
Page 114
!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!
!!
Thesis and Dissertation Services