Algal Turf Scrubber (ATS)

Algae to Energy Project

Cleaning Rivers while Producing Biofuels

and
Agricultural and Health Products

Report to:
David Orr and Adam Lewis
The Lewis Foundation
% The Lipson Group Inc.
1422 Euclid Avenue, Suite 1500
Cleveland, OH 44115
January 31, 2010

Walter H. Adey
Department of Botany
National Museum of Natural History
Smithsonian Institution
Washington, D.C. 20560
Adeyw@si.edu

1
 Table of Contents

Executive Summary

Introduction

The ATS Energy Project

The Algae
Primary Production
3-D Solution to the Diatom problem
The ATS Ecosystem
Nutrient Removal
Oxygen Injection
By-Products of ATS
The All-Terrain ATS
Issues to Consider

The Economics of ATS Operation

Plans for Project Expansion

References

Appendix
A. Algae of the Test Systems, Smithsonian Institution
B. Susquehanna Project, University of Maryland
C. Ozark Highlands Project, University of Arkansas
D. Biomass Processing, Western Michigan University
E. Project Extension (Chesapeake Algae Project, ChAP/Chesapeake Algae
Consortium): organizations listed above, plus
Blackrock Energy, StatoilHydro, Exelon Power, Constellation
Energy, HydroMentia, Ecological Systems Technology
University of William and Mary/VIMS,

2
 EXECUTIVE SUMMARY

Developed initially to control aquatic microcosms, Algal Turf Scrubbing (ATS)

has a 30-year history of integration with the chemical and biological function of
living ecosystems. Scaled-up, over the last 15 years, to several acre dimensions
for tertiary sewage treatment, aquaculture and remediation of eutrophic creeks
and canals, it has been used, coast to coast, in the southern U.S. The purpose of
this investigation, funded by the Lewis Foundation, was: (1) to expand ATS
technology to river scale, especially in more northerly climes; (2) to develop by-
products from ATS-produced algal biomass, including biofuels; and (3) to develop
the economics of large watershed application. To accomplish these tasks, several
test ATS floways have been established on rivers, and numerous working
associations with university scientific teams and companies have been developed.
The major finding of the field experimentation of this study on large non-
point-source waters was a shift of ATS community structure from a standing crop
dominance of filamentous green algae to that of filamentous diatoms. The
biodiversity of the ATS algal communities was thereby greatly increased,
enhancing the basic rationale of ATS. However, the lesser shear strength of
diatom filaments, in the moderate energy environment of ATS, increased slough
rate, and reduced expected productivity by 25-50% and nutrient removal by 10-
30%. This issue has been resolved with the development of a new type of basal,
3-D screen that retains diatoms, more than doubles algal productivity, and
produces an entirely new field of potential process improvement. While broader
use of this new technique is necessary, it is likely that further enhancement of
nutrient remediation is now possible. Interface Global has joined our team, and is
working to achieve mass production of these new screens; it seems likely that
cost increases will be minimal as compared to process enhancement.
Photosynthesis produces oxygen, and the experiments of this study have
emphasized that oxygen production by ATS is large; by manipulation of ATS flow
to prevent excess supersaturation, injection of more than 35 tons/ATS acre/year
into source waters should be possible. The more serious implication of water
body eutrophication is hypoxia (dead zones), and in watershed amelioration,
more emphasis needs to be placed on direct removal of hypoxia. This study
produces an initial plan to accomplish this task for Chesapeake Bay.

3
 Aliquots of the algal biomass produced on the ATS floways of this study
have been shipped to our associate team of chemists at Western Michigan
University, where they have developed a physical/chemical separation process for
carbohydrates, oils, proteins and minerals. The processed carbohydrate solutions
were forwarded to our associate chemical engineering team at the University of
Arkansas, where they have been converted to hydrogen and butanol using a
membrane-modified, double Clostridium fermentation. The W. Michigan team
has also demonstrated omega-3, PUFAs in the ATS algal oils. While it seems likely
that omega-3 separation will be more economic than the transesterification to
biodiesel process that the Western Michigan team has also demonstrated, this
should be determined at pilot scale. In plant growth tests, the residual proteins
and minerals have been shown to be a quality fertilizer. However, the large
quantity of silica diatom frustules, with a very large surface/volume ratio, in the
mineral-rich fertilizer may additionally provide soil remediation that can lead to
significant capturing and storing of atmospheric carbon. Both chemistry teams
are now ready to begin expansion of this refinery process to pilot scale.
Previous large scale ATS systems have required time-consuming
construction, earth-moving, grading and surface-preparation methods that have
provided limitations to scale up. In this study, a new, at par cost, All-Terrain ATS
unit has been developed that provides for rapid construction, dimensional
flexibility, and applicability to unstable and difficult terrain such as landfills and
river bottomland. This system will be used extensively in planned watershed
scale-up.
With the development of an array of by-products, and the ability to work
quickly and at large scale on rivers, tributaries and bays, this project has
presented a preliminary economic plan for ATS utilization that can reduce the cost
of both water remediation and algal by-product production. In the under way
process of scale-up to pilot, and then full scale, for the Chesapeake Watershed, it
will be necessary to further enlist environmental economists and modelers to
guide and verify watershed-scale processes. However, managed by Blackrock
Energy, and with the direct support of scientists from five universities, and
companies such as Exelon Power, StatOil Hydro, Constellation Energy,
HydroMentia and Ecological Systems Technology, this next step of The
Chesapeake Algae Project, as the Chesapeake Consortium, seems well within
reach.

4
 INTRODUCTION
Algal Turf Scrubbing (ATS) is a system for utilizing algal photosynthesis to control a
wide variety of water quality parameters. Developed in the early 1980s at the Smithsonian
Institution, as a biomimicry of coral reef primary productivity, ATS was initially used as a tool to
manage an extensive series of living microcosm and mesocosm models of wild ecosystems. The
models ranged from coral reefs to estuaries and fresh water streams; one of the largest of the
mesocoms was a Chesapeake Bay system grading from tidal fresh waters to full coastal
salinities. Applied to closed living models, ATS functioned to control nutrients, oxygen levels,
carbonate systems, including calcification (through CO2 control) and to minimize toxic
compounds from the local human-engineered environment. ATS also allowed the development
of planktonic communities and planktonic borne reproduction in model ecosystems, as it has
little effect on the planktonic component. This early development of ATS and the models it
controlled are described by Adey and Loveland (2007). Those authors also describe how ATS
techniques, scaled up, can restore our damaged rivers, bays and lakes.

Successfully scaled-up for nutrient removal, from point-source and semi point-source
open waters, during the 1990s and early 21st century, ATS use ranged from aquaculture and
tertiary treatment of sewage to agricultural canal amelioration of nutrients. By 2009, eight
scaled-up ATS systems had been built and operated from coast to coast, mostly in the southern
tier of states. The northernmost unit was constructed on the lower eastern shore of Maryland.

Okeechobee, Florida, 2003 7-acre ATS system

Tilapia farm
2.5 acre Algal Turf Scrubber (ATS) Falls City, Texas
processing 10-20 Mgpd of farm stream water

Harvesting algae

5
 Two particularly successful early ATS operations were a 7 acre Tilapia operation in Falls
City, Texas which produced commercial quantities of fish for nine years, and a acre, 250 K
gpd tertiary sewage system in the northern Central Valley of California (Craggs et al, 1996). In
the earlier ATS floways, considerable algal primary production was achieved (yearly means
from 25-45 g(dry wgt)/m2/day), and in some cases the algal biomass was used as an animal feed
and fertilizer. In these ATS systems, the algal source was local, mostly by self-seeding from
source water, and the algal biodiversity was high, across the spectrum of algal groups.
Filamentous green algae were dominant, in terms of produced biomass and nutrient uptake.
During the last decade, the Ocala, Florida engineering firm HydroMentia built a variety of
several acre dimension ATS systems for State/municipal water quality control of small non-
point-source waters (agricultural canals, creeks), but had restricted its activities to Florida. A
multiple-site 1440 acre ATS system had been designed to remove agricultural nutrients from the
Suwannee River in northwestern Florida (Stewart, 2006), but at this time that project has not
been funded beyond engineering design phase or implemented by the state government.

The purpose of this study, funded by a generous gift from the Lewis Foundation to the
Smithsonian Institution, and described in this report, was four-fold: (1) to demonstrate non-point-
source nutrient capture capability by ATS in stream to river scale environments with mostly
lower nutrient concentrations than point-source waters; (2) to develop a system for expanding the
utilization of produced algal biomass to biofuels and other by-products; (3) to develop a process
for expanding the use of ATS capabilities to watershed scale, especially in the Chesapeake Bay
Region, currently under both a Federal Executive Order and a Federal Court Order to solve
nutrient/hypoxia problems; and (4) to demonstrate an economic basis for large-scale ATS non-
point-source treatment, with a complex of by-products, in more northerly latitudes.

To accomplish these tasks and acquire the necessary associates and supporters, numerous
ATS power point presentations were made at universities, NGOs and state and Federal
government agencies throughout the Chesapeake Bay Watershed and at associated universities
outside of the Watershed. Three university teams headed by Dr. Patrick Kangas at the
University of Maryland, Dr. Marty Matlock at the University of Arkansas and Dr. John Miller at
University of Western Michigan were inducted into the study; these scientific teams were

6
provided Smithsonian subcontracts to demonstrate ATS capabilities, and to examine refining
processes for the ATS algal biomass. The final reports of these partners are appended.

Associations were also formed with scientific teams at the University of William and
Mary and its Virginia Institute of Marine Science (VIMS) on the York River on southern
Chesapeake Bay, the Chemistry Department at the University of Arkansas (a butanol
fermentation group headed by Prof. Jamie Hestekin), and several companies: Exelon Power,
Blackrock Energy, HydroMentia, Statoil Hydro, Ecological Systems Technology, Constellation
Energy and Interface Global. The contributions of these groups are discussed in this report along
with several process diagrams to document their participation and the results of joint work.
However, since they were not subcontract participants, funded by the Lewis Foundation gift,
reports were not requested. Blackrock Energy, having brought together a coalition of additional
universities and companies, is now serving as the lead organization in this consortium.

ATS floways were established on the Susquehanna, Great Wicomico and York Rivers in
the Chesapeake Watershed and on the Springdale tributary of the Illinois River in Arkansas.
Algal biomass from these systems and several of HydroMentia ATS units in Florida were sent to
the University of W. Michigan for the study of its chemical composition and the development of
separation processes. The extracted carbohydrates were sent to the Department of Chemical
Engineering at the University of Arkansas and were processed to butanol and hydrogen using a
semi-permeable membrane modification of the Ramey, Clostridium double-fermentation.

Muddy Run, Susquehanna River - Three ATS floways VIMS, York River
7
 THE ATS ENERGY PROJECT

The Algae

Previous work with ATS, from microcosm to moderate scale had concentrated on
systems rich in either hard benthic environments (i.e. rock or branches), with abundant algal turfs
as biofilms, or aquatic flowering plants that carried periphyton on their stems. Those ATS
systems were highly dominated by filamentous green algae, particularly species of the genera
Cladophora, Spirogyra, Microspora, Ulothrix and Rhizoclonium. Diatoms and cynobacteria
were usually present, especially as epiphytes, but rarely provided significant biomass.

Typical filamentous, green dominated algal turf, growing on a basal plastic screen from a Florida
ATS. Diatoms in the canopy in the diagram at the left; will form a brown tint later in cycle at right.

This investigation concentrated on river-stream systems, where either because of their

large volume to bottom ratio and/or the predominance of sandy or muddy bottoms, planktonic
algal communities, and those diatom communities specialized to living on sand and mud bottoms
(rather than periphyton and algal turf communities), were dominant. While extensive seeding
efforts from small, rocky periphyton-dominated local streams, with abundant filamentous green
algae, were attempted, in most test floways, for most of the operational time, planktonic-sourced
diatoms came to provide the majority of the biomass. The dominant species, by biomass, of
8
these ATS floways were filamentous diatoms, particularly species of the genera Melosira,
Fragilaria, Diatoma and Berkeleya. In species number, however, the unicellar diatoms, species
of Nitzschia, Gomphonema and Navicula, were prominent. Investigations of the biodiversity and
community structure of the algae on the floways of this study were carried out, and these species
and genera, along with their abundances at each site are shown in Appendix I. In general, algal
species biodiversities on these ATS systems were very high and this was especially true of the
Susquehanna floways, where over 200 species of algae were tallied.

Berkeleya rutilans

Melosira nummoloides
Several diatoms
cf. Synedra sp.

9
 While some green algal species, filamentous and unicellular, were present on all ATS
systems of this study, they were mostly minor elements of biomass. The filamentous green alga
Spirogyra was moderately abundant on the Susquehanna floways during the summer, and the
tube-former Ulva (Enteromorpha) was ubiquitous on the floways of estuarine rivers.
Nevertheless, over time, we learned that the overwhelming biomass dominance of diatoms could
not be significantly altered, in spite of considerable effort during the first six months, particularly
at the Susquehanna site, to seed other species of periphyton. Similar efforts were undertaken for
a shorter period in Arkansas. This was not necessarily a productivity concern, as diatoms,
especially in temperate to arctic coastal waters, are the dominant producers of the plankton.
Unfortunately, due to basic physiological and structural differences between diatoms and green
algae, diatom filaments lack the tensile strength to consistently remain attached in the moderate
energy environment of ATS systems, and this is a requirement of normal ATS function. Unlike
in the earlier green algal-dominated ATS systems, significant slough, off the floway surfaces,
occurred throughout the growing cycle (typically 7 days, though reduced to 5 days in mid-
summer and increased to 14 days in mid winter).

ATS VIMS
May-July 2009 Cyanobacteria
2%
Chlorophyta
3%
Dinophyta
0%

Berkeleya rutilans (43%)

Melosira nummoloides (26%)
Licmorpha sp complex (7%) Bacillariophyta
Gyrosigma sp. (5%) 95%
Nitzschia sp. complex (5%)
Fragilariopsis sp. (3%)
Pseudonitzschia cf. multiseries (2%)

York River ATS: 95% of biomass is diatoms (Bacillariophyta

10
 On the Muddy Run (Susquehanna) floways, algal slough was captured at the end of the
floways with mesh bags, and was shown to be 10-20% of the harvested biomass. However, the
mesh size of the bags had to be large (30um) to prevent blockage and flooding. Diatom
filaments, made up of weakly attached cells, were prone to breakage, and subsequent passage of
the smaller individual cells through the slough bags. Also, algae sloughed to the mesh bag ceased
to efficiently photosynthesize, and add to system productivity. In addition, some algae likely
died and were degraded by bacterial action, between harvests, in the slough mesh bags.
Although ATS-function was otherwise normal, all of these processes reduced harvest. On the
other hand, nutrient removal was only minimally effected, and this aspect of ATS operation will
be discussed below. After several months operation on the Susquehanna floway at 2% slope, a
second floway of 1% was added, side-by-side with the first unit in the hopes of reducing diatom
slough. This second, lower-sloped floway had little effect on diatom dominance or retention,
and algal productivity was lower. On the other hand, nutrient removal increased. This is probably
a result of longer retention time, and greater organic particulate removal with the lower flow rate.

Muddy Run, Susquehanna River, Floways

Wood and Aluminium Floways species group comparison

2% slope (aluminum) & 1% slope (wood)

70 (wood)

60

50
Number of Species

40

30

20

10 Wood

Aluminium

0 Total
E

E
IA

E
E

A
EA

EA
EA

EA

EA
EA

YT
R
TE

PH
YC

YC
YC

YC

YC
YC
C

IO

PH
PH

PH

PH
PH
BA

AP

O
O

O
O
O

LA
EM

YD
R

YS

ID
N
N

LO

LE

IL
YA

H
R
N

M
C

AP
G
H

C
G

LA
C

BA
EU
C

ZY

R
H
C

11
Primary Production

The high photosynthetic capability and primary production of algae is well known and
has been the subject of the development of a potential new industry for algae to biofuels (Chisti,
2007; Ryan, 2009). However, most of the algae utilized in these new systems, whether
photobioreactors or high-rate ponds are suspended, (floating or planktonic) in the water column;
optimizing and harvesting that photosynthetic potential, economically, has proven difficult. ATS
was developed as a simple, low cost means of achieving and utilizing algal photosynthetic and
productivity potential using attached algae. When nutrients are moderately high and solar energy
moderately abundant, the results of ATS are striking, and this has been repeatedly demonstrated
in the southern tier of states from coast to coast. Numerous aquaculture and point-source or
semi-point source treatment projects with productivities ranging from 25-45g(dry wgt)/m2/day,
have been developed (Adey and Loveland, 2007; Mulbry et al, 2008; www.Hydromentia.com).
While these rates of biomass productivity may not be as high as some photobioreactor methods,
they are 5-10 items that of local agriculture. When used to clean wastewaters, as well as develop
biomass products, the cost of the products, such as biofuels can be considerably lower than
photobioreactor methods. Conversely, when ATS is used to produce algal by-products, the costs
of nutrient removal from eutrophic waters can likewise be significantly reduced.

The ATS productivity initially achieved in this project, especially on the Susquehanna
River floways (Table 1), where relatively low nutrients and annual solar energy were factors,
were less than expected. While still about 3-4 times greater than local agriculture, test ATS
units produced at 20-50% less than that of the seasonal expected values. As discussed above, this
was largely a result of the formation of a diatom-dominated community that tended to slough-off
the floway, and in the case of greatest reduction that we further describe below, chironomid
activity. The solution to this loss of productivity has been a relatively minor, low cost
tweaking of the ATS process, in part by increasing micro surface area in the growing screen.
These new substrate screens, further discussed below, are expected to generally double
production and nutrient removal on diatom-dominated systems and thereby bring non-point
source treatment at least to the levels routinely achieved in point source amelioration.

12
 Table 1
ATS Floway Comparison
summer ( 6-9 )
High nutrients ------------------------------------this study------------------------------ Low nutrients

Site Patterson, CA1 Springdale, AR Susquehanna, PA Susquehanna, PA S-154/Suwannee, FL.2

(type) (tertiary) (large sub urban creek) 2% slope-alumn. 1% slope-wood (river)
length area 153.8m 1100m2 92.3m 30m2 92.3m 30m2 92.3m 30m2 184.6m 5.76.106m2
flow rate 644l/min 77/l/min 50 l/min 44 l/m 12 Bl/day
mean Summer Solar 590 546 (72.103) (63.4 103) 540
(Gates, 1980)
Productivity 40 30.9 17.6 12.3 45
gdry)/m2/day
Algal Composition Nutrient Removal Incoming Nutrients

(mg/l)TN 5.0 9.68 1.38 1.38 1.08

(g/m2/d)NLd) 4.5 36 3:31 22:04 2.25
N/P 1.6 :1 62:1 11.5:1 11.5:1 10.1:1
mg/l TP 3.1 0.157 0.12 0.12 0.107
g/m2/d) PLd. 2.81 0.58 0.288 0.254 0.223

%TN 22 30.5 28% 37% 44%

gN/m2/day 1.11 10.99 0.93 1.08 1.00

%TP 45 52.9 9% 17% 83%

gP/m2/day 0.73 0.31 0.026 0.043 0.185

% diatoms 23 76 77.5 77.5 (30)3

% filament. Greens 69 20 14.5 14.5 (65.5)
% cyanobacteria 7 4 6 6 (2)

1. Craggs et al, 1996 2. Stewart, 2006 3. Adey et al, 1993

13
3-D solution to the diatom problem

About half-way through this investigation, it became clear that on larger muddier rivers,
including estuarine rivers, a diatom community dominated by species of the filamentous
genera Melosira, Berkeleya, Diatoma and Fragilaria would self-establish. Repeated seeding of
periphytic green algae from smaller tributaries would alter this situation significantly only for a
few weeks. The highly dominant diatom community of the rivers included filamentous species
that could enter the benthic community of the ATS and swamp normal benthic species because
of the overwhelming abundance of their cells and filaments in the continuously overflowing
water.

As demonstrated earlier, the green-algae-dominated ATS community operating on a

Florida canal, shows that similar diatom elements that occur on river ATS are also present (in
that case different species of Melosira and Eunotia), but they do not displace the dominant green
filaments. In those cases, the diatoms are epiphytic on the green filaments and become an
important (up to 30%), but not over-riding element of the algal turf. Visually this is seen as a
browning of the growing biomass as it nears harvest, and the diatom filaments fill in the upper
structure of the algal forest.

The traditional substrate in the ATS system is a plastic screen. Many varieties of plastic
screen have been employed, although a HDPE plastic of 3x5 mm mesh is typical, a wide range
of mesh size has been used. On the scale of the ATS floway and the enhanced algal community
these screens are 2-dimensional structures. The dominant diatom communities that occurred on
river ATS systems quickly attach to these standard screens, but their filaments constantly
shear-off in the moderate energy environment of an ATS, producing a lower standing crop and
ultimately lower water remediation capabilities and by-product biomass. Diatom filaments have
an entirely different structure from that of typical green algal filaments. In the latter case, the
often massive cellulosic wall is continuous from cell to cell, usually with no break. Individual
cells can die without compromising the integrity of the green filament, at least in the short term.

Diatoms, on the other hand, are basically cellular in construction, and have an entirely
different kind of cell wall. The young, naked, diatom cell develops vesicles in its plasmolemma

14
membrane, which in turn secrete amorphous silica within a matrix of protein, polysaccharide and
lipid. When the rather complex silica frustule of typically four parts, two valves and two thin
encircling girdles (holding the valves together) is completed, the entire complex is encased in a
matrix of polysaccharide (sulphonated glucoromannan) (van den Hoek, 1995). The silica unit is
the frustule; it is long-lasting, and after the death of the cell, it can become fossilized, sometimes
as extensive deposits (diatomaceous earth). The frustules, and their individual cells are glued
together by polysaccharides, often at spines, or held together in a polysaccharide matrix that can
allow cell to cell sliding. Diatom unicells are often mobile, and can slide with a conveyor belt
like movement of the matrix along grooves in the frustules. In some genera, such as Berkeleya,
the filament is an extension of the polysaccharide sheath, in which the individual cells are
randomly arrayed. Diatom cells can quickly attach to a substrate with the polysaccharide glue
of their wall, and that is why they generally the first colonizers of new surfaces.

In conclusion, unlike green algal filaments (or red or brown algal filaments in sea water),
diatom filaments are basically fragile and subject to breakage in the energy-rich ATS
environment. To prevent this loss of diatom biomass, it was decided to develop a 3-dimensional
screen to simulate the filamentous green algal structure that we could not maintain. Since in
earlier systems, diatoms that attached to and entangled in the canopy of the dominant
filamentous green algae remained on the floway, it seemed logical that if we supplied that
structure artificially, we could more efficiently retain the diatoms on ATS floways.

A wide variety of off-the-shelf, deep pile throw rugs, with 1-2 cm thick loose fibers, were
established on an experimental ATS system on the Chesapeakes Great Wicomico River (a 10-14
ppt environment at about mid-Bay, north to south). Also, several screens were specially woven,
with 1-2 cm long Dacron fibers, to simulate the structure of the green alga forest that we could
not consistently develop on river ATS units. The Dacron was employed because it would
provide for minimal degradation under solar UV. Some of the rugs had a structure that
considerably enhanced diatom retention. However, after several months of experimentation, the
specially woven Dacron 3-dimensional (3-D) test screens provided the greatest productivity of
all the test screens. These were established in the central part of the ATS test floway. Standard
2-dimensional (2-D) ATS screens were arrayed both above and below.

15
 Regular Screen Fully Developed Raised Filament

Comparison of 2-D and 3-D screens prior to harvest

The 3-D screens consistently provided over 2 times the productivity of the standard 2-D
screens (see below). Note that the plotted points are means of two 3-D and three 2-D screens; a
t-test provides a high level of significance, p=0.0003. The fitted curves are sine curves anchored
to expected peaks in June and December. This study is still underway, and as of late December,
2009, shows a production on 3-D screens of 2.7 x 2-D screens. The mean production for 3-D
screens during December, the month of least incoming solar radiation, was 11.5 g/m2/day, as
compared to 4.0 g/m2/day for the standard 2-D screen; this was twice the winter production on
the Patterson tertiary ATS (Craggs et al, 1996) and about the same rate of production as standard
2-D screens in Florida during December (Adey et al, 1993).

2-D screens 3-D screens

ssssscreens
screesssssssc
reensscreens
16
 50.0
45.0
40.0 Great Wicomico River
35.0
30.0 2D vs 3-D Screens 2-D
25.0 3-D
20.0 2D Sine
15.0 3D Sine
10.0
5.0
0.0
8/15/09 9/4/09 9/24/09 10/14/09 11/3/09 11/23/09 12/13/09 1/2/10

With the 3-D screen returning the structural environment of the standard complex, mini-
forest algal turf, filament-sloughing is largely avoided and the expected productivity, based on
previous small-stream, tertiary treatment and aquacultural systems, in the Chesapeake Bay area
is more than recovered. It is likely that additional improvements in screen design will further
increase filament retention and productivity. Interface Global carpet Company of La Grange,
Georgia has been working with us to produce new, entirely synthetic, rug-like screens that are
similar to our successful 3-D units. While several more iterations are probably necessary, it
seems likely that the needed screen can be produced at large scale and low cost. Note that
vacuum harvest is used in all of these studies, and the resulting biomass has been investigated to
be certain that no synthetic fibers are present that might artificially increase harvest biomass.

There are likely many factors that are at work in the 3-D process. A major factor in
productivity increase in photobioreactors is increase of surface area to limit solar self-shading by
planktonic cells. That is a macro-approach that was seen in our early coral reef work (Adey and
Steneck, 1985). In part, this current methodology is a micro approach of increasing surface
area to provide additional attachment sites for diatom filaments. However, the 3-D process also
creates cells or mini-volumes, a habitat in which water pulsing is likely modified to a swirling
turbulence that increases mixing without filament shearing.

17
The ATS Ecosystem

Most algal to by-products production systems, including those producing biofuels, are
monocultures of single algal species, and in many cases, the organisms employed are
genetically-selected varieties. Genetically modified clones or GMOs (genetically modified
organisms), have also been employed, and it is often suggested that the algal to biofuel industry
will be successful only using species that have been modified to have specialized characteristics,
such as oil production (Ryan, 2009). In production modes, these cultures are not bacteria-free,
but they are at least as much monocultures as most agriculture crops, and require constant
vigilance to prevent algal weeds and disease from overcoming the target algal species.

In a genetic context, an ATS floway is highly diverse with wild algal species. It is
considerably more than a polyculture of several human-selected species, and perhaps is best
termed a multi-culture of whatever algal species are available in the water body being treated. It
is a culture, because the physical conditions are established by human action, the species best
adapted to those conditions settle (as any wild ecosystem), and routine harvest takes place;
inevitably some species are selected (by the conditions created) and others at least minimized.
Only a single pair of ATS floways have been intensively monitored for algal diversity (the units
at Muddy Run, on the Susquehanna River), and over 200 species of algae have been tabulated.

The typical ATS community additionally contains a diverse mixture of protozoans and
small invertebrates, especially nematode worms and insects in fresh water and small crustaceans
and polychaetes in estuarine waters. Such an ATS floway is an autotrophically-based ecosystem.
However, by examining the percentage of nutrients removed as TN (total nitrogen) versus DN
(dissolved nitrogen), the floway nutrient sink can be demonstrated to be typically 30-40%
derived from particulates in the water column (including plankton). Thus, the ATS ecosystem
clearly has a heterotrophic component. As earlier microcosm research had shown (Adey and
Loveland, 2007), particulates are captured not generally by the filtering of plankton, but rather
mostly by capture of organic particulates. Some of the protozoans and invertebrates on the ATS
floway could be capable of mucous web capture of particulates, but two of the most abundant
organisms on ATS systems, diatoms and blue-green algae, secrete abundant mucous to encase
their cells, attach themselves to the substrate and , in the case of the diatoms, provide mobility.
18
 Most diatoms can function as both autotrophs and heterotrophs (based in dissolved
organics), and some can switch from one source to another (Graham and Wilcox, 2000).
Although it has yet to be investigated, it is likely that there is an internal transfer of nutrients and
energy from organic particulates through the bacteria, protozoan and invertebrate sub-
community to diatoms, and this may be partly the source of apparently higher levels of diatom
photosynthesis on ATS systems.

Nutrient Removal

Table 1 sums up the status of nutrient removal in the three primary floways of this study
(see Appendix for full data). These are compared with previously published work, at high
nutrient levels for the Patterson, CA tertiary treatment system (Craggs et al, 1996) and for the
HydroMentia S-154 (2.5 acre) system in Florida, using as a proxy the engineered application to
the Suwannee River (HydroMentia, 2005). The warm season data (June-Sept.) was employed
because data were available for all compared systems for that time. Mid-winter data was absent
due to winter system failure on the Susquehanna River; Springdale Creek (Arkansas) had a
broken data stream and is being re-run during 2009-2010. The table is set up with the three
diatom-dominated floways in the middle, the moderately high nutrient Patterson system on the
left and the low nutrient Suwannee (S-154) on the right. The Patterson and Suwannee (S-154)
floways are green algal dominated.

The extensive work of Mulbry and his associates (Mulbry et al, 2008) using ATS on
agriculture wastes was considered for comparison in this table, since many of their conclusions
support the earlier work at acre scales mentioned above. However, the Mulbry systems were
recirculating, and low volumes made multiple passes over the ATS surface, whereas typical ATS
remediation is single pass and involves large daily volumes. As a result, the Mulbry systems
show high removal to loading ratios, although considering loading levels, nutrient removal is
well within the range for single pass ATS. Although recirculating ATS is quite applicable to
highly concentrated agricultural wastes, it would not be suitable for non-point source treatment.
Where appropriate, this body of work is cited and discussed below.

19
 The basic engineering design of ATS systems provides for a small, pulsing (or surging)
water flow, on a very shallow floway, moving against the algal filaments which are held tight to
growing screen on the floor of the floway. This low energy, oscillating water motion limits the
capability of algal cells to deplete nutrients in the film of water immediately adjacent to each
cell, thereby increasing nutrient uptake potential and primary productivity. Light flashing, due to
alternate cell to cell shading and lighting in the pulsing water, also provides for a greater
photosynthetic rate. The resulting higher photosynthesis and production rate (over natural
periphyton and micro-algae situations) produces greater nutrient uptake rates both by
metabolic (direct uptake) mechanisms and precipitation within cell walls for some nutrients,
(e.g., P, Ca, Mg, Fe).

To some extent, at moderate to low concentrations, nutrient removal rates in ATS

systems are a function of nutrient concentrations in the body of water being treated. Mulbry et al
(2008) and the engineering firm HydroMentia (www.HydroMentia.com) have demonstrated a
direct relationship between nutrient loading (concentration X flow/unit area) and nutrient uptake
for total nitrogen (TN) and total phosphorus (TP). At least at low to moderate concentrations of
TN & TP, that uptake is directly proportional to nutrient loading. However, this relationship
does not extend to higher loadings where nutrients can become saturating. Mulbry et al
(2008) have shown that nitrogen saturates in ATS algae at about 6.5% and phosphorus at about
0.9% (at low to moderate pH). In addition, earlier studies at very high nutrient levels showed
that cyanobacteria can become dominating. Cyanobacteria are photosynthetic. However, many
species can form extensive biofilms, and often are embedded in thick mucilage that limits the
mixing effects of the ATS system that act to break down the stagnant water adjacent to the
algal cells. Thus, potential productivity and nutrient removal can be significantly reduced at high
levels of loading (Adey and Loveland, 2007). Also, at low nutrient concentrations, algae have
considerable ability to produce nutrient depleted tissues (Adey, 1987). This is especially the case
in ATS systems when nutrients in influent waters are unbalanced from the ideal N:P of
approximately 16:1. Further complicating the generalized relationship between nutrient loading
and uptake, algae have the capability to store nutrients when they are available in excess of
growth requirements (Mulbry et al, 2008). Finally, as shown in this report, restructuring the
basal substrate (2D vs 3D), can radically change productivity and potential removal. Thus, at
20
this time, there is no global, fine-tuning model for nutrient removal in all ATS systems,
although a specific system can generally be characterized within narrow parameters.

Available solar energy is the ultimate limiter of primary production in ATS systems, and
where yearly data are available, a sine curve of both primary biomass production and nutrient
removal can be produced (Craggs et al, 1996). Algal growth occurs throughout the 24-hour day
and metabolic nutrient removal (eg, nitrogen) is independent of daily solar radiation. On the
other hand, phosphorus removal (as well as Ca, Mg, Fe, etc.) can be dependent upon solar
radiation/photosynthesis removal of CO2 , followed by elevation of pH and the resultant
precipitation within algal cell walls. Where pH levels can rise high enough, as in the example
shown below for the Springdale Creek ATS system, such precipitation is directly proportional to
solar energy capture on a diurnal basis. Algal community structure, in the physical sense, as
demonstrated above, is also a determinant of primary production (and therefore nutrient
removal). Nevertheless, in spite of wide day to day variability in any one system, removal
follows productivity and tends to be consistent, varying seasonally. Given access to local wild
communities of algae, algal production and nutrient removal in ATS systems is not significantly
temperature dependent, as long as freezing of the water does not occur. This is consistently seen
in production curves that follow light, even when temperature lags by several months.

10.5 Springdale Arkansas Floway - pH

10
9.5
9
pH

8.5
8
7.5
7
8/19/2009 8/20/2009 8/21/2009 8/22/2009 8/23/2009 8/24/2009 8/25/2009
Outflow Inflow

Returning to Table 1 and several of the ATS systems utilized in this project, all three
diatom rich sites are lower in biomass productivity than would be expected based on incoming

21
summer solar energy alone and the production rates achieved in Florida and California. The
Springdale, Arkansas site is about 25% lower and the Chesapeake sites are about 40% lower.
However, it is quite likely that utilization of 3-D screens on these systems, as demonstrated
above, will achieve expected productivity or higher. The Springdale floway, can be compared to
Patterson since it has a relatively high nitrogen loading. Here at 77% of the productivity of
Patterson, the system produces a total nitrogen (TN) removal to load of 30.5%, 8.5% higher
than Pattersons For total phosphorus (TP), the removal to load ratio is 53%, 8% higher than
Pattersons. Thus, the utilization of 3-D screens to double biomass productivity is likely to make
the Springdale floway one of the best ATS nutrient removal systems yet constructed.

In comparing the Susquehanna floways to that for the S-154/Suwannee design, summer
production at the Susquehanna site was 50-60% less than that in southern Florida, even though
in-coming summer solar radiation is about the same. Nevertheless, the total nitrogen removal
rates are about the same as Suwannee and the removal to load ratio is about 25% lower.
Doubling production with 3-D screening should bring Susquehanna summer removal rate
considerably higher than Suwannee. For phosphorus, the removal rates are about 20% of
Suwannee and the removal to load is 9 and 17% respectively of the Suwannee design. Doubling
production with 3-D screens would still make this a poorer performer at its current length,
although higher pH levels, with higher productivity, would improve performance. However, the
Suwannee floways at 188m are twice the length of the Susquehanna floways, and the additional
length is likely to significantly increase phosphorus removal due to precipitation. The mean
available P in the summer Susquehanna River water is sufficient to provide a P removal rate that
is slightly better than the Suwannee removal rate; thus the Susquehanna floway at twice the
productivity (due to 3-D screens) and twice the length would likely be a superior performer for P
removal. Although increasing system length is likely to reduce the efficiency of N removal, with
doubled productivity it is still likely to be better than either Patterson or Suwannee.

All of these high diatom-dominated floways would have to be operated with 3-D
substrate to be certain of performance potential. However it seems likely, given the information
available, that diatoms on an ATS system can provide better nutrient removal per unit biomass
than an ATS community that is dominantly green algae. Perhaps the diatom capability of

22
removing dissolved organics is a factor. More research is necessary, as greater understanding
may further improve performance. Do the generally smaller diatom cells provide greater surface
area for cell/water exchange in the turbulent environment of ATS? Do the silica frustules of
diatoms become a lens or diffuser that improves light utilization in the light-flashing
environment of ATS? Initially perceived as a problem, the dominance of diatoms on river scale
ATS may open up considerably greater potential for water remediation and by-product
development. Also, while conventional wisdom would maintain that nutrient removal is best
done at or near critical point sources, this is not necessarily the case if management
considerations and optimization of refinery production of by-products are considered. In
addition, low to moderate concentration of nutrients are proportionally better utilized by ATS
systems than higher concentrations.

Oxygen

Hypoxia as a result of excess nutrient-created algal blooms becomes a major factor of

water quality degradation in rivers, lakes, and bays and even on very large bodies of water such
as the Gulf of Mexico. Point source, bacterial-based nutrient treatment methods significantly
aggravate this problem, especially if costly oxygen (air) injection is not widely employed. ATS
injects a significant amount of pure oxygen into water bodies being treated. The amount of
oxygen theoretically being injected is somewhat more than the tonnage of the dry organic
biomass (ash-free) being produced, and even though it is not given a dollar value in traditional
remediation, it could be the most important single factor in early stage clean-up. It can only be
available using photosynthesis-based rather than heterotrophic bacterial-based waste-water
treatment.

Diurnal studies of oxygen production were undertaken on both the Susquehanna and the
Springdale ATS floways of this investigation. An example, over several days, from the
Springdale study is shown below, followed by a summary of warm season diurnal oxygen
production on the 2% (aluminum) floway on the Susquehanna River. The mean warm season
collected algal production on this latter floway was 17.6 g(dry)/m2/day, and theoretically net
oxygen production should be about 30% higher than organic (ash-free dry weight). Thus, with
organic production (ash-free dry weight) about 53% of dry weight (see By-Products, below), net
23
production of oxygen should exceed 13 g(dry)/m2/day, although this analysis shows only 8.12
gO2/m2/day. However, when oxygen is allowed to supersaturate by several hundred percent, as
in this study, oxygen is being lost to the atmosphere; in this case, more than one third the amount
being produced is lost. This is not surprising, since routinely, on a sunny afternoon, in the algal
turf on an ATS floway, abundant oxygen bubbles can be seen. These bubbles tend to lift the algal
filaments up into the flow, being responsible, to some extent, for the tendency of the diatoms to
break free and join the slough. When oxygen injection is important, increased water flow can
avoid high levels of supersaturation and deliver maximum oxygen to the water column.

300 Oxygen Saturation State, Springdale, Arkansas

250
DO Saturation (%)

200
150
100
50
0
8/19/2009 8/20/2009 8/21/2009 8/22/2009 8/23/2009 8/24/2009 8/25/2009
Outflow Inflow

Modeling studies relating ATS function to that of the water body being treated, in
conjunction with pilot operation, are essential to optimizing the results of ATS treatment.
Hypoxic zones generally occur in deeper water and tend to be locked in by overlying warm
water in summer, although co-modeling can likely direct the solution of this problem.

Mean Daily Net Production

8.12 gO2/m2/day

24
By-products of ATS

During the 1990s and the early years of the 21st century, the algal biomass produced in
ATS operation had been widely used (although largely experimentally) as an animal feed (fish,
cattle and chickens) and as a fertilizer or soil amendment. Walter Mulbry and his team at USDA
studied the role of ATS algal biomass as a slow release fertilizer; they also developed an
effective system for ATS in-farm cleaning of waste nutrients from cattle waste, with the algae
being used as a field fertilizer for corn (Mulbry et al, 2005).

With the increasing recognition that many valuable products could be obtained from algal
culture, it became clear that the economics of water clean-up, using ATS, could be significantly
improved with further by-product development. During the 1980s and 90s, the possibility of
algal biofuels was examined by the DOE, National Renewable Energy Laboratory (Sheehan et
al, 1998). Since that time many algal to energy projects have developed. Some algal to biofuels
projects have concentrated on the production of biodiesel from oil-rich monocultures of
planktonic or microalgae following an extensive search for strains with high levels of oil storage
(Chisti, 2007). Much of this work relies on photobioreactors, and could only with difficulty be
adapted to waste water treatment. Some effort has been undertaken to genetically modify algal
strains, using recombinant DNA to combine desirable characteristics from one or more algal
strains into the ideal species. This has the same basic problems of escape to the wild and
dilution of natural biodiversity as agricultural modification. However, with algal to biofuel
methods, such a process would be pushing the envelope at both ends of the scale, very large area
and microorganisms.

A very different approach, using high rate ponds (raceways with circulating paddle-
wheels), converts algal biomass to methane gas, and has been applied to polluted farm run-off in
southern California. Also, as noted above for diatoms, some algae have the capability to ingest
both soluble and particulate organics, heterotrophically, independent of their normal mode of
photosynthesis. However, this requires an independent source of organics, presumably waste
biomass. A recent review of the field by the NRDC, Cultivating green Energy, The Promise of
Algae Biofuels (Ryan, 2009) provides a critical examination of the processes currently
underway to convert algal biomass to energy. Since the result of most algal production is a
25
biomass with high water content, an energy/carbohydrate conversion requiring drying before
refining makes little life-cycle sense. Thus, fermentation is the most likely first step in a
conversion process.

The production of ethanol by yeast fermentation of plant sugars has become the dominant
liquid biofuel in the U.S. However, ethanol carries low energy per liter (73% of gasoline), is a
strong solvent, and is highly volatile, characteristics requiring special handling. It is processed
as a low percentage mixture with gasoline. An alternate fermentation, ABE, using Clostridium
bacteria, was developed during the first World War, mostly for acetone, the butanol component
serving as an industrial solvent. ABE was the primary source of butanol as an industrial solvent
until mid 20th century, when conversion from crude oil refining began. The four carbon alcohol
butanol, can function as a direct 1:1 replacement for gasoline, and has 96% of the energy per liter
of gasoline. It is also six times less evaporative than ethanol, is less corrosive, less susceptible to
water contamination, and produces less CO2 per BTU of energy supplied..

This century-old fermentation, using a variety of carbohydrates, had been greatly

improved during the 1990s by David Ramey (Ramey and Shang-Tian Yang, 2004;
www.butanol.com) using two separate, sequential cultures of Clostridium tyrobutyricum and
Clostridium acetobutylicum with hydrogen and butyric acid produced in the first stage and
butanol, from the butyric acid, in the second. Nevertheless, the process had remained somewhat
unstable, and with an improved but still modest efficiency.

More recently, Dr. Jamie Hestekin, a semi-permeable membrane specialist, had established
a laboratory, at the University of Arkansas, funded to use membranes to improve the Ramey
butanol fermentation. Since his R&D process was nearing successful completion, using corn
sugars, the decision was made in this project to concentrate on butanol as the algal biofuel, and
to look for other by-products from the relatively small quantities of residual oils. The primary
question we had to answer was: would the algal sugars, and other associated carbohydrates, be
different enough from the standard corn sugars to upset the Clostridium tyrobutyricum in the
front stage of the modified Ramey process?

26
 The Ramey Process, a dual, two-
species, Clostridium fermentation

Modified two-step Ramey process

utilized at the University of Arkansas

27
 Algal biomass from all of the primary ATS test sites, except Springdale, Arkansas, and
from several of the HydroMentia ATS units in Florida, especially an estuarine unit on the
Caloosahatchee River draining to the Gulf of Mexico Coast, was used to develop a bench scale
separation process. Dr. John Miller, and his associates at the University of W. Michigan, first
devised a pattern of physical/chemical separation for the diatom-rich, but complex algal biomass
from the ATS units, as shown in the following diagram. Although some of the carbohydrate
solution was processed to ethanol, through several procedures as described in the University of
W. Michigan report in the appendix, as an economic process this approach remains inconclusive
at this time.

Algal Composition & Biofuel products

Butanol /
Ethanol

Co-Products
High N, P Biodiesel
High Silica
PUFAs

9 Nov 2009 13

Several batches of the algal-derived carbohydrate solutions were forwarded to the

Hestekin Laboratory at the University of Arkansas and were successfully processed to hydrogen
and butanol. The advantage of the Clostridium fermentation in the Hestekin process is its general
lack of sensitivity. We now know that the bacteria can accept both soluble algal sugars as well
as the insoluble carbohydrate components. Although basically anaerobic, the fermentation is not
upset by small quantities of oxygen, making it relatively easy to operate as an industrial process.

28
 According to the estimates by the W. Michigan Laboratory, 36 gallons of butanol per/dry
ton of algae would be produced from the ATS algal feedstock. At approximately 30g(dry
wgt.)/m2/day (48 tons/acre of ATS/year), as discussed above, this would yield 1700 gallons of
butanol per acre per year. If all of the oils present in the algal biomass were to be converted into
biodiesel, a relatively simple transesterification process, also described in the W. Michigan
report, an additional 800 gallons of biodiesel could be produced. However, approximately 20%
of total oils are composed of high value Poly Unsaturated Fatty Acids, (omega-3 oils). As
marketed retail today omega-3 oils are typically at a concentration of about 25%, and it seems
likely that a mature processing operation would produce either high value Omega-3 or biodiesel
fuel. In the economic analysis that follows, omega-3 production is shown as the more likely
process. The residual oils excess to omega-3 production could be transesterified to biodiesel or
converted to some other products. Omega 3s are currently obtained from keystone fish that are
not being sustainably harvested, so this direct algal source could be environmentally crucial.

Algal Biomass Pre-

processing
Percentages indicate
relative amounts of dry algal
biomass after each step

9 Nov 2009 14

29
 In the NREL investigation of micro algae for biofuels (Sheehan et al, 1998), the target was
oils, particularly for transesterification to biodiesel. Some oil-rich diatom species were included
in that work. However, as seen in the work of John Miller at al. in this study, most diatoms have
low oil content. On the other hand, as noted above, diatom cell walls are rich in easily-reduced
polysaccharides that appear tol be acceptable to Clostridium tyrobutyricum without prior enzyme
digestion. Cellulose appeared in these algal turf assays, but may have derived from the associated
green algal filaments; the algal turf biochemistry of this study included some samples from
Florida ATS units in which green algal filaments were likely the dominant algae. Initially seen as
valuable in monoculture for oil-based biofuels, natural diatom cultures are apparently best-suited
for butanol fermentation, with the oils having modest omega-3 content of high value for health
food processing.

Finally, roughly 22 (dry) tons per acre per year of nutrients, minerals and residual
proteins are left after the extraction of the higher value by-products. As the W. Michigan study
has demonstrated (see Appendix D), and as shown by Mulbry et al (2005), this can be used as a
high quality fertilizer. As shown below, the abundant silica diatom frustules are highly
ornamented, and possess abundant pores or areolae. Diatom frustules are typically on the scales
of 10s of a um in dimension, and their spines, ridges and areolae on a um scale. However, it has
also been demonstrated that the frustules have a porous and cellular nanostructure at the nm
scale that provides virtually infinite silica surface (Crawford et al, 2009). If properly processed
so that the silica diatom frustules remain, the very high surface area of this ash component
would likely provide a soil amendment that increases tilth, water and ion retention. This in turn,
under an appropriate cultivation routine, could provide increased organic buildup, improved
agricultural production and, at large scale, removal of significant carbon from the atmosphere.
Soil studies are necessary to determine whether this route to carbon sequestration is viable.

The pre-processing procedure, as shown here, would probably have to be modified to

bring forward the polysaccharide enzymes, perhaps with mild ultrasound treatment, before
mechanical chopping. The silica frustules being relatively dense could be removed with
centrifugation and then the organic residue treated with ultrasound sufficient to break down the
30
plasmolemma and release cells contents for continued processing. At the end of the refining
process, the silica frustules could be re-mixed with the residual nutrients, minerals and proteins
to produce the high grade fertilizer discussed above. Breaking down cell walls to get at the cell
contents is a typically difficult early step of producing a biofuel. It is especially difficult in
higher plants where the cellulose is usually infiltrated with the highly resistant lignin. In diatoms,
the silica cell wall is embedded in polysaccharides, relatively easily degraded and usable
compounds.

Residue of a portion of algal turf from a Muddy Run floway. Treated to remove organic materials
and examined under the Scanning Electron Microscope (SEM), the abundant ornamentation and
virtually infinite surface area of the complex silica frustules of diatoms, is apparent. Kept intact in
processing, these frustules will greatly increase the fine porosity of soils in which the ATS refining
residue of proteins, nutrients and minerals is employed as a fertilizer.

31
The All-Terrain ATS

The Algal Turf Scrubber process requires modest land area and often is criticized for this
characteristic. However, as Table 2 shows, 15 times as much biofuel per unit area can be
produced from ATS than from corn. At least in agricultural regions where corn is grown for
ethanol, or soy beans for biodiesel, it makes little sense to argue that ATS requires too much
land. Later in this report, an ATS model for Chesapeake Bay will be described which will
provide quite sufficient land for Bay recovery. Nevertheless, there are situations in
urban/suburban regions where land would be difficult to obtain. Also, there are otherwise ideal
sites along rivers in rural areas where instability of substrate or irregularity of terrain would
make standard ATS units impractical. While small temporary test ATS units can be established
to discover the basic parameters that can lead to an efficient full scale system, this can be very
time consuming. The larger scale units developed and utilized over the last decade require
extensive grading and surface preparation as well as considerable concrete construction. Once
committed, it is very difficult to alter these systems to new or undiscovered conditions.

Algae vs. Corn Biofuel potential

ATS Algae Corn
Biomass yield 50 3.1 dry ton/acre/yr
% Carbohydrate 25% 35%
Carbohydrates available 12.5 1.1 ton/acre/yr
Fermentation

% fermented 90% 90%

Carbohydrates used 11.3 1.0 ton/acre/yr
Fermentation efficiency (C into fuel) 51% 51% =malcohol/(malcohol+M 2C O2)
Fuel produced 5.75 0.5 ton/acre/yr
potential ethanol production 1822 159 gal/acre/year

potential butanol production 1775 155 gal/acre/year

triglycerides 6%
Biodiesel

Triglycerides available 3 ton/acre/yr

conversion efficiency 95%
Triglycerides used 2.9 ton/acre/yr
Potential Biodiesel production 810 gal/acre/year

Algae Yields are estimates from Dr. Walter Adey.

Algal carbohydrate compositions from Dr. John Miller (WMU) measurements of ATS samples
Corn yields are from the US Dept. of Energy Billion-Ton Vision.)

32
 The All-Terrain ATS was developed specifically to utilize the otherwise waste area of
capped landfills. However, because of the ease and rapidity of All-Terrain ATS construction, its
flexibility for expansion in width or length, the potential for rapid re-siting, at cost at par with
standard ATS units, this system opens up land areas that would otherwise be unavailable for
ATS water remediation. An additional value of the All-Terrain systems is that it allows for the
efficient injection of stack gas CO2 both to improve productivity as well as provide for additional
CO2 removal from the atmosphere. While Mulbry et al (2008) were not able to demonstrate CO2
enhancement of productivity and nutrient removal in their re-circulating systems, their dairy
water nutrients were added one time in the morning and diurnal effects were not studied. It is
likely that nutrient availability and carbon availability did not coincide in the experiments. Since
CO2 addition has been routinely used successfully in photobioreactor production (Chisti, 2007),
it is possible that this was a experimental flaw in the Mulbry et al study.

All-Terrain ATS of one unit newly installed on the York River at VIMS
33
Issues to Consider

The ATS process of producing algae for water quality remediation, while providing an
array of bi-products, including biofuels, has been remarkably free of significant problems; a few
require mention.

The ATS process has been developed, in part tuned to avoiding diseases and the very few
invertebrate pests (grazers) which could inhabit biomass production and nutrient removal.
Especially on river systems, where the algal biodiversity of ATS units is particularly high,
significant disease is highly unlikely. Viral or bacterial invasion would likely just shift
community structure, with unaffected species taking over dominance. On the other hand, a
limited array of algal invertebrate grazers can be present. Amphipods are small, shrimp-like
arthropods that have many types of feeding behaviors. Some of the commonest species in
benthic environments are grazers of algae. Data taken on the VIMS floway shows that biomass
build-up and grazing amphipod abundance are closely related. When this relationship is
considered, and harvest schedules closely adhered to, amphipod grazing has never been a
significant issue in marine or estuarine ATS systems.

VIMS Algal Floway: Optimal harvest schedule

Emmett Duffy and Elizabeth Canuel, VIMS
Algal accumulation
Experiment
Plots harvested daily for 10 days
measured algal production, grazers

Daily production Results

Algae produce rapidly
2 4 6
Pests invade more slowly
Optimal harvest at ~6 days (summer)

Grazer invasion

1 October2009 ChAP Project Kick-off meeting

34
 In freshwaters, amphipod grazing is replaced by a single group of insects, the Midges.
The non-biting, non-eating flying stages are reproductive phases that lay eggs in shallow water
and on wet surfaces. The resulting larvae, called chironomids, are grazers of algae that grow in
those waters and are potentially goat-like in greatly reducing algal productivity potential.
During the past 20 years, in 14 out-door ATS floways in which chironomids could have been a
problem, they have had a known presence of note in only four units. In two of those (the
Patterson and Springdale systems) they occurred only periodically and were easily controlled by
close observation and early harvests. At one tertiary treatment unit on the eastern shore of
Maryland, where the ATS unit was unfortunately located adjacent to a sewage plant skimmer
with a continuous production of Midges, chironomid infestation was serious. This case was
controlled by Bacillus thuringiensis a bacterial addition inimical to the larvae; as a result,
successful operation was achieved.

During the second summer of operation of the Muddy Run Susquehanna floways,
chironomid infestation was serious (virtually none appeared the first summer). While this
reduced biomass production considerably, it apparently had minimum effect on nutrient removal.
Presumably, the chironomid larvae consumed a considerable amount of algae, reducing biomass
(potentially available energy as biofuel), but retained a large portion of the nutrients which were
removed with the larvae on harvest. Unfortunately, this site was remote from the University of
Maryland base of operations, and adequate short harvest methods could not be routinely
applied. This situation was also highly unique in that the larvae apparently arrived with the
water flow from the stored water reservoir. The stored water reservoir emptied daily, producing
large areas of mud flats, perhaps uniquely suited for both Midge egg-laying and larval release.
That the infestation occurred in one year and not another, might be related to weather conditions.
In general, chironomid grazing in ATS freshwater systems can be viewed as a potentially
periodic, but usually a quite controllable issue. It has never been a problem on the larger Florida
ATS units or on the large aquaculture unit in south Texas. On the other hand, those systems tend
to be characterized by flocks of small, wading birds, which likely feed on any invertebrates
present in the ATS. No large ATS system, of an acre or more, has suffered significant
chironomid infection.

35
 Considering the typical mode of grazing used by chironomid larvae, it seems likely that 3-
D screens will reduce their effectiveness. The larvae construct small mucous/detritus tunnels
from which they foray out to graze. It is unlikely that this can be accomplished on the raised
filaments, and, as a result, they would be restricted to the basal part of the screen. This issue
needs to be examined further, but it is likely that a screen can be devised that will exclude
chironomid larvae from effective grazing.

A single floway in Florida has been struck by periodic algaecides; these were likely
herbicides used on orange groves and flushed into the creek being treated during heavy rains.
While potentially a serious problem in treating small creeks, dilution would likely prevent this
problem on large rivers. Nevertheless, the poisoning of algae at tributary scale should be
regarded as a more serious problem than eutrophication, since algae are at the base of most
aquatic food chains. If algae cannot be grown due to anthropogenic toxins, then we have created
dead water swimming pools out of our natural water. In this regard, ATS systems can be
regarded as a water canary, capable of warning that waters are unacceptably toxic. ATS can
function in this way also by degradation or accumulation of low level industrial and agricultural
toxins.

Two of the ATS systems in this project (Susquehanna and Springdale) were shut down by
freezing for about two months each. Freezing of the water in these systems occurred, in spite of
considerable efforts to keep them running. Neither the Great Wicomico nor the York River
raised floways, at mid and lower Chesapeake Bay respectively, were seriously affected by the
hard and prolonged cold-wave of early January, 2010. Up to 40oN latitude on each coast and
35oN in mid country, properly designed and operated systems should be able to overcome this
problem. Nevertheless, as briefly described above, the co-location of large scale ATS with fossil
fuel power plants can allow ice-free winter operation (using waste heated water effluent from the
power plants) as well as to provide for significant CO2 recycling.

36
 THE ECONOMICS OF ATS OPERATION

Since well before the Green Revolution, the potential for algal culture systems to produce
valuable products has not been in doubt. However, the cost to do so, in competition with
agricultural products, most of which are grown on land cleared from forest centuries ago, has
been a serious obstacle to the advancement of algal systems. Unfortunately, the full costs of
agriculture, including soil and nutrient loss to natural waterways, are seldom included in
agricultural economic analyses, so that a true comparison is difficult. In addition, in parallel with
human experience over millennia of agricultural development, monocultural algal solutions, and
more recently genetically modified solutions, have been sought. There is little question that this
can be done, since algal isolates have been cultured in the laboratory for many decades.
However, the complexity of achieving agricultural levels and methods of algal production over
decades, when the processes for higher plants took millennia, provides a serious economic test.

The ATS system has overcome most of the large scale algal culture problems by using
algal weeds that are available, year-round in the waters being treated. Moreover in moving to
non-point-source treatment, it has become clear that an enormous diversity of weedy species is
readily available in natural environments, and disease and competition are not issues of concern.
Pests and excessive chemical pollution provide minor concerns, as we noted above. However,
the economics of constructed systems as opposed to open field agriculture remains the primary
issue in the production of algal products. As shown below, if value is achieved for water clean-
up, ATS provides a very cogent economic scenario.

Table 3 presents an economic model for treatment of the Susquehanna River at or near its
confluence with Chesapeake Bay. The 3000 acre model is employed because at this dimension
ATS will remove the amount of phosphorus independently assessed as necessary to return the
upper Bay to nutrient health; at this dimension, ATS will also inject enough oxygen into the
water column to securely remove the hypoxia (dead) zone that has come to characterize the
upper Bay. In addition, an engineering design, with economic analysis, for the Suwannee River
in Florida has been completed by Hydromentia and provides the basic accounting to develop
such a large scale model.

37
 Chesapeake Bay Journal, November, 2009
Karl Blankenship. Report released for state of
Cheasapeake Bay and its rivers
Required to Reach Target:
Chesapeake Bay P reduction 1.45Mlbs/year
Susquehanna River P reduction 1.04 Mlbs/year

3000 acres of ATS on Susquehanna Effluent

will remove: 1.5Mlbs of P; 9Mlbs of N/yr
will potentially add: over 200 Mlbs of oxygen/yr
will potentially increase upper Bay oxygen
concentration by 7 mg/l, thus removing hypoxia

Currently efforts are underway to join the stack gas CO2 of a Constellation coal-fired
power plant, as well as warm water effluent to ease winter operation, with the more-than-
adequate area potentially available at the Aberdeen Proving Ground, at the confluence of the
Susquehanna River and the Chesapeake Bay. This base is presently being reviewed for base
closure. Whether or not this action is completed, there is a large area on this base that could be
considered for co-use. The economic model presented in Table 3, is based on this situation,
although other sites are also possible.

As the economic model demonstrates, given a nutrient trading or banking scenario, cost
is not a significant factor in Bay clean-up. Indeed, it is likely that utilized in this manner, ATS
clean-up of Chesapeake Bay is a green economic engine. Similar operations could be developed
on the Potomac and James Rivers, potentially using available military land. This would not only
provide a return of Chesapeake Bay to health, but would provide an economic boost to the Bay
in greatly improving the fisheries of these waters that have been so long under environmental
stress.

Table 3 provides an economic analysis based on a set of ATS units of 3000 acres total
near the mouth of the Susquehanna River. Costs are derived from the Suwannee River (1440
acres) engineering design, with a 20% discount based on new technology and scale up factors.
Nutrient values are Chesapeake Bay Commission (CBC) published mean nutrient removal costs
38
from 2004. Based on current anecdotal information this provides about a 27% discount on the
current values (approximately $200/kg P $10/kgN). However, as we shall show, allowing for the
value of by-products, ATS systems operated by a public, non-profit entity solely to reduce
nutrient removal costs, could remove nutrients (and carbon) at about 50% of current costs.

The analysis of Table 3 is written to be carried out by a quasi government agency, with
30-year construction bonds at 5%. Carried out by a private engineering firm the difference
between ATS nutrient removal costs and the market price, minus the cost of capital, would have
to be negotiated. Given a functioning nutrient banking system, a significant return on investment
could be available to drive clean-up efforts.

39
 Table 3
Economic Analysis
Upper Chesapeake Watershed Project
3000 Acres Based on Suwannee River ATS Design
ATS Production: 30g(dry m2/day (48 tons/ac/yr)

Banked Credits for Nutrient Removal (Biomass Prod. 144,000 tons/year)

(Chesapeake Bay Comm., 2004; discount approx. 27% current cost)
Nitrogen (3%) 4320 tons @ $3.86/lb. (CBC $8.50/kg) $33.4M
Phosphorus (0.4%) 576 tons @ $66 /lb. ( $145.20/kg) 76.0M
Carbon (38%) 54,720 tons @ $20/ton 1.1 M
$110.5M

Sale Returns on By-Products

Biofuel (butanol) 36 gals./dry ton, 5.2M gals. @$2.50/gal $13.0M
Fertilizer (est. 54% of biomass) 77.8K tons @ $200/ton 15.6M
Hydrogen 2 M lbs @ $2.10/lb (energy equiv.) 4.1M
Omega-3 oils, 3.4 gals./dry ton, 490K gals @$100 gals. whsl. 49.0M
(Biodiesel, optional 2.5M gals. At $2.50/gal; 15.6M)
$ 81.7M
Gross Return: $188.1M

Costs

3000 acre ATS, upper Chesapeake Bay

Capital (5%) 30 yrs. plus O&M , incl. land Suwan. Proj. 06 $127 M
Refining (all products, est. 20% of product value) $15.5M
Administration (10% of capital cost + O&M cost) $12.7M
Profit or Re-investment $32.9M

Gross Cost: $188.1 M

40
 The refining process for by-products needs to be expanded to pilot scale and then
financially re-analyzed. Also, life cycle analysis needs to be carried out, especially for the
refining process. It seems likely that the hydrogen produced in butanol fermentation would be
most efficiently utilized on site to provide power for refinery operation. Also, CO2 injection into
ATS systems has not been performed at large scale, although the technology exists to do so using
the All Terrain system, and application does not appear to be difficult. This aspect of project
economics would have to be linked to carbon trading. Rivers in the Chesapeake Watershed have
moderate alkalinity and in-water CO2 supply is not a serious problem. The considerable rise in
the pH of test systems shows that CO2 stress is being placed on productivity. CO2 injection,
spread-out on floways can significantly increase algal productivity. On the other hand, this will
delay phosphorus precipitation down the floway. These two opposing needs will have to be
worked out in pilot operations. ATS has already been constructed and routinely operated for
several years at multi-acre scale at a half-dozen sites. It is certainly time to expand to a 25-50
acre dimension pilot with a similarly scaled-up refining plant.

41
 PLANS FOR PROJECT EXPANSION

Now that it has been established that ATS remediation (nutrient removal and oxygen
injection) is technically and economically feasible for non-point source waters (rivers and bays)
at least up to mid latitudes, it is essential that this endeavor expand to multi-acre pilot
demonstration scale. Since the Chesapeake Bay has demonstrated significant environmental
problems, and the Federal Government has committed itself to solving this long-standing issue, it
is here that our efforts will be concentrated.

A coalition of interested parties, convened and managed by Blackrock Energy, including

the partner universities of this study plus the Universities of Virginia and Penn State University
and the energy companies Exelon and Constellation Energy, have indicated a strong interest in
forming a consortium to move ATS to pilot scale. This coalition has developed a project for both
water-cleaning and by-products, including butanol, on the upper Chesapeake Bay. The
Constellation coal-fired plant adjacent to the Armys Aberdeen Proving Ground will provide the
land, heated effluent water and stack gases from its plant to expand to the next phase. Scientists
at the Universities of W. Michigan and Arkansas are prepared to ramp up over one year, first to
algae shipped by the barrel and then to a mobile refinery hauled to the Constellation site by
years end.

Exelon and Constellation plan to work together to support this endeavor with the goal of
sourcing both funding and the necessary political support. Once fully successful at pilot scale (1-
3 acres), the intent is to expand to a 25-50 acre industrial pilot and within a few years to full
acreage necessary for upper Bay clean-up. Tentatively, land area on the Aberdeen Proving
Ground is being considered for this scale up. Additional efforts will also be initiated to develop
similar facilities on the Potomac and James Rivers, thus looking to ameliorate the entire Bay as
well as to provide significant quantities of by-products, including biofuels. A series of meetings
are planned by all parties during early 2010 to bring this effort to fruition.

A recent proposal, developed by Blackrock Energy to the U. S. Department of Energy, in

support of the Algae Consortium effort, is included in the Appendix. Constellation Energy and
Exelon Corporation are the lead organizations in this proposal.

42
 REFERENCES

Adey, W. 1987. Food production in low-nutrient seas. BioScience 37: 340-348.

Adey, W. and K. Loveland. 2007. Dynamic Aquaria: Building and Restoring Living ecosystems.

3rd edition. Elsevier, Amsterdam. 508 pp.

Adey, W, C. Luckett, and K. Jensen. 1993. Phosphorus removal from natural waters using

controlled algal production. Restoration Ecology 1: 29 39.

Adey, W. and R. Steneck. 1985. Highly productive eastern Caribbean reefs; synergistic effects

of biological, chemical, physical and geological factors. NOAA Symposium Ser. For

Undersea Research 3: 163-187.

Chisti, Y. 2007. Biodiesel from microalgae. Biotechnology Advances 25: 294-306.

Craggs, R. W. Adey, B. Jessup, and W. Oswald. 1996. A controlled stream mesocosm for the

tertiary treatment of sewage. Ecological Engineering 6 : 149-169.

Crawford, S., A Chiovitti, J. Picket-heaps, and R. Wetherbee. 2009. Micromorphogenesis during

diatom wall formation produces siliceous nanostructures with different properties.

J. Phycology 45: 1353-1362.

Gates, D. 1980. Biophysical Ecology. Springer-Verlag, Berlin. 611 pp.

Graham,L. and L. Wilcox. 2000. Algae. Prentice-Hall, Upper Saddle River, 640 pp.

Mulbry, W., S. Kondrad, C. Pizarro and E. Kebede-Westhead. 2008. Treatment of dairy manure

Effluent using freshwater algae: algal productivity and recovery of manure nutrients using
pilot-scale algal turf scrubbers. Bioresource technology 99: 8137-8142.

Mulbry, W., E. Kebede-Westhead, C. Pizarro and L. Sikora. 2005. Recycling of manure

nutrients: use of algal biomass from dairy manure as a slow release fertilizer.

Bioresource Technology 96: 451-458.

Ramey, D. and Shang-Tian, Yang. 2004. Production of butyric acid and butanol from biomass.

Final report to DOE for contract DE-F-G02-OOER86106.

43
Ryan, C. 2009. Cultivating Clean Energy the promise of algae, NRDC and Terrapin Bright

Green, LLC. pp1-81.

Sheehan, J. T. Dunahay, J. Benneman and P. Roessler. 1998. A look back at the U.S. Department

of Energys aquatic species program: biodiesel from algae. Golden, Colorado, National

renewable Energy Institute. NREL/TP 580 24190.

Stewart, E. 2006. Preliminary engineering assessment for a comprehensive algal turf scrubber-

based nutrient control program for the Suwannee River in Florida. 19pp + 18 sheets.

Van den Hoek, C. 1995. Class Bacillariophyceae. (In) Mann, D. and J. Jahns. Algae, An

Introduction to Phycology. Cambridge University press.

44
 Appendix to:

ATS to Energy Project

A. Algae of the ATS Test Systems, Smithsonian Institution

B. Susquehanna Project, University of Maryland
C. Ozark Highlands Project, University of Arkansas
D. Biomass Processing, Western Michigan University
E. Project Extension (Chesapeake Algae Project, ChAP/Chesapeake Algae
Consortium): organizations listed above, plus
Blackrock Energy, StatoilHydro, Exelon Power, Constellation
Energy, HydroMentia, Ecological Systems Technology
University of William and Mary/VIMS,

45
 Appendix A
The ATS Energy Project

Algae of ATS Test Systems

Haywood Dail Laughinghouse IV

Department of Botany
National Museum of Natural History
Smithsonian Institution
Washington, DC 20560
and
Department of Environmental Science and Technology
University of Maryland
College Park, MD 20742

46
Background

The word algae was first used by Linnaeus in 1753 in Species Plantarum to denote a
group of cryptogams (plants with hidden reproduction, without flowers or cones). However, the
organisms that are englobed by this term do not constitute a singular evolutionary taxonomic
group and are polyphyletic including both prokaryotic and eukaryotic nucleate organisms, within
many distant clusters. A common characteristic within all the algae is that they contain the green
pigment chlorophyll a within their cells, providing photosynthesis and the ability to maintain an
autotrophic life, although some algae can also be heterotrophic during parts of their lives. Some
algal groups, including the diatoms that figure prominently on the ATS systems of this project,
also have accessory pigments (brown, yellow, blue, and red), capable of capturing solar energy
and providing the coloration of the group that is other than green.

Algae are a highly diverse group of organisms, and can be found in an array of different
ecosystems. Most of the large macroscopic species live in marine waters (e.g. kelps and rock
weeds), while the microscopic species have a wider distribution, though these microscopic,
basically single-celled species, can accumulate in colonial and filamentous types and can form
macroscopic mats or algal turfs. Many species occur in fresh and brackish waters, though they
can also be found in snow, thermal vents, humid soils, desert crusts, in/on plants, animals, rocks,
etc. Basically, we live in an algal world.

In the aquatic environment algae are mostly planktonic or benthic. Most benthic
(periphytic) algae are either blue-green algae (Cyanobacteria), green algae (Chlorophyta),
diatoms (Bacillariophyta), or red algae (Rhodophyta), though on the freshwater ATS systems of
this study no rhodophytes were observed, however the red algal genus Polysiphonia was
encountered on the estuarine systems, especially that of Great Wicomico River, which has
around 14 salt.

The colonization by algae of an aquatic surface is a linear process of the settlement rate
directed by the size and type of the source system (if this is dominated by bedrock, SAVs,
sediment, etc.), substratum texture, light intensity, and water velocity (see Biggs, 1996). In our
cases, the source waters favored by the ATS system, tends to be dominated by diatoms (Class
Bacillariophyceae) (see Figures below obtained by microscope counting). These are important
components of many algal communities. In natural short-term communities, a succession
usually occurs where first an organic matrix with bacteria will develop on a substrate, with small
adnate diatoms, then larger colonial diatoms, and so on up to larger filamentous greens
(Chlorophytes). Since there is a weekly harvest on ATS floways, we tend to keep the system at
the intermittent stage of colonial diatoms. Since the primary objective in ATS systems is to
maximize algal growth rate, and therefore nutrient removal, leaving algae on the floway too long
will cause sloughing and loss in biomass (Biggs, 1996).

47
 Diatoms are unicellular organisms that can sometimes form colonies which are
filamentous, radial, or spiral. Their bivalve cell wall (resembling petri dish), is characterized by
having a polysaccharide cell wall containing silica (frustule). Both valves are surrounded by
girdle bands, which also contain silica. In certain species, like the tube-forming Berkeleya
rutilans, numerous frustules are found within a thick polysaccharide mucilage/sheath. Besides
containing chlorophylls a and c, diatom plastids have the brown to yellow pigments fucoxanthin,
diadinoxanthin, and diatoxanthin, allowing for the golden-brown color of the Bacillariophyceae.
The group stores its reserve principally as chrysolaminarin (glucose 1-3 ramified in 1-6), but
the cells can also accumulate lipids (Reviers, 2006), hence its large use in many studies for
biodiesel production. Diatoms are capable of moving (gliding) on substrates by excreting
mucilage through their raphes (fissure running the length of the valve of some pennate diatoms).
This mucilage is involved in their adhesion to the substrate, as well. Another interesting feature
is that some diatom species are capable of burying themselves under the sediment at times,
controlled by an internal biological clock (Hoek, 1995).

On the left, Luticola cf. geoppertiana from Muddy Run ATS seen through SEM, demonstrating
some of the characters used for morphological analysis. On the right, Cymbella tumida with many
Navicula seen using LM and phase contrast.

Diatom frustules are extremely ornamented, some visible in the light microscope (LM),
like raphe, striae, and nodules, but others only visible under scanning electron micrscopy (SEM),
like pores, areolae, channels, and diverse processes like (rimoportulae, fultoportulae, spines)
which are found internally or externally in the valves. The presence, location, disposition, and
number of these structures are the characters used to identify the genera and species of the
Baciallariophyceae (Barber & Haworth, 1994), since these are different in evolutionary lineages.

ATS floways also contain chlorophytes, a.k.a green algae. This group of algae
differentiates from the diatoms by containing chlorophylls a and b, as principal photosynthetic
pigments, though in some cases the accessory pigment siphonoxanthin or lutein are found (Sze,
1993). Like the higher plants that green algae gave rise to, chlorophytes present a green
48
coloration since, in general, they do not contain other pigments masking the color. This group
usually uses starch as a reserve. The morphology of this group is diverse, containing unicellular,
colonial, coenobial, flagellated, filamentous, (pseudo)parenchymatous, etc. species. In the algal
turfs, besides metaphytic chloroccocalean ceonobial species like Pediastrum, Scenedesmus, and
Desmodesmus, true filamentous forms were found such as Ulothrix, Cladophora, Stigeoclonium,
Oedogonium, and Rhizoclonium.

Oedogonium is a common green algae of relatively calm freshwater habitats, which was
true in the ATS floways of Muddy Run, from the middle to the bottom. The filament (thallus) is
unbranched and attached by a holdfast. Besides interesting morphologically characters, like
striations or caps in the cells from cell divisions, reticulate chloroplasts, and large vacuoles,
among others, this alga has internal thickening of its cell walls following cell divisions. This
ring consists of noncellulosic polysaccharides (hemicellulose) (Sze, 1993), and cellulose is
deposited on the inside of this ring.

Photomicrograph of green algal clump from Muddy Run ATS. a)

Cladophora glomerata with epiphytes (diatom Cocconeis), b)
Oedogonium sp. whose wall doesnt allow the attachment of
epiphytes.
Numerous Cladophora spp. were seeded on the ATS floways, though great long term
success was not achieved. This alga is branched and also attaches by a holdfast. B. Whitton
(pers. comm.) comments that for Cladophora to grow well in an environment, both high
nutrients and high light intensity, is required while other authors comment that there is a
seasonality effect to Cladophora, where late winter/early spring is often diatom dominated,
progressing with cyanobacteria (e.g. Phormidium sp.) in the summer and patchy growth of large
filamentous algae, like Cladophora in late summer (see Biggs, 1996). A seasonal cycle
resembling this was seen on the Muddy Run ATS floways and the Springdale floway increased
to 35% of their relative biomass in green algae towards the end of the summer season. Each cell
49
of Cladophora is surrounded by a thick cellulosic cell wall which is continuous throughout the
filament; it lacks a mucilaginous sheath, allowing for the growth of epiphytes. This is different
than Oedogonium or Spirogyra that contain a sheath that does not permit epiphytic growth.

Cyanobacteria were also found on the floways, though usually presenting less than 5% of
total biomass. The most common species were of the Phormidium complex. This group, by its
name, is a prokaryotic alga, possessing chlorophyll a and photosystems I and II, carrying out
photosynthesis in the presence of oxygen. They possess phycobilisomes with accessory
pigments such as phycoerythrin, phycocyanin, and allophycocyanin (Reviers, 2006). These
produce a typical blue-green color for the group. Morphologically, the group is simple,
difficulting precise identification of many species, although the vegatitive structure can be
coccoid, colonial or filamentous, and contain mucilaginous sheaths. There are various types of
storage in the cyanobacteria: nitrogen reserves are constituted of cyanophycin and carbon
reserves formed by cyanophycean starch and lipids, you can also find polyphosphate granules in
the cells. Their cell walls are composed of peptidoglycan, characteristic of eubacteria. Of
ecological importance is the capability of cyanobacteria to metabolize atmospheric nitrogen,
particularly when nitrogen concentrations in the water column are very low; this is facilitated by
the presence of heterocytes, although species without this cell are also capable of using
atmospheric nitrogen.

Phormidium sp. from periphyton collected at Muddy

Run ATS
Sampling

Throughout the study period, samples of periphyton were collected from all of the ATS
Systems of this project and subsequently analyzed at the National Museum of Natural History
(NMNH). For qualitative analyses, periphyton was scraped off the screen with a spoon at
various parts of the floway to have an understanding of the full complexity of the algal
community. For quantitative analyses, top, middle, and bottom portions of the system of a

50
known size (between 50-100 cm2) were scraped and sent to NMNH for analysis or portions of
the screen were sent and then processed in the same manner.

Quantitative analysis

Samples of thick algal mats were homogenized (blended) in distilled water, where the
final volume was recorded. A subsample of 100ml was taken and standard lugols solution was
added to sediment and stabilize the samples for analysis preparing it for counts in the inverted
microscope. After about 3 hours, or enough time for lugol to penetrate the algal cells in the
subsample, the sample was homogenized by shaking slowly back-and forth for 1 minute, and
5ml was placed in a sedimentation chamber of the same volume. This was let sit for a minimum
of 6 hours, so that the algal cells could sediment to the bottom of the chamber, allowing
visualization in the Zeiss inverted microscope for counting.

Algal counts of periphyton were conducted on a Zeiss inverted microscope, counting a

minimum of 500 cells/valves. These were recorded and a formula proposed by Wetzel & Likens
(1979) modified by Schwarzbold (1992) was used to determine the number of algal cells p/cm2:

N = n . V / v . 1/S

where:
N = number of individuals per cm2
n = total number of individuals counted in the sample
V = volume of the sample with scraped material in ml
v = volume of the counted fields
S = surface of the substrate in cm2

Qualitative analysis

Diatoms were cleaned in preparation from LM and SEM, this is a different procedure
than just visualization that is conducted for soft algae (green and blue-green algae). Twenty ml
of sample were removed and boiled in hydrogen peroxide 30% in a l:1 sample ratio, adding
potassium dichromate during the process to speed oxidization, until returning to the original
volume of 20 ml. The cleaned material was then concentrated in a centrifuge, and rinsed several
times until reaching a neutral pH (c. 10x). This material was diluted into 3 different dilutions, air
dried on coverslips, and processed for either LM or SEM analysis.

For LM, the coverslips were let dry overnight, and then mounted onto glass slides with
the high infraction index (1.704) mounting medium Cargille Meltmount . The samples were
then analyzed under 400x or 1000x (oil immersion) using an Olympus Bx50 compound
microscope with phase contrast. Morphological characteristics were recorded for the taxa for
specific delimitation, i.e. valve/cell diameter and length, diameter of central area, striae length,
density, and number.

51
 For SEM, coverslips were mounted on specimen stubs and sputter-coated with a
gold/palladium alloy (10nm) and analyzed using a Leica Stereoscan 440 SEM.

The algae were qualitatively analyzed according to specific literature for each group:
Bacillariophyta (Krammer & Lange-Bertalot, 1986, 1988, 1991, 1991; Cox, 1996; Round et al.,
1990); Cyanobacteria (Komarek & Anagnostidis, 1998, 2005); Chlorophyta (Prescott, 1982;
Bicudo & Menezes, 2005; John & Williamson, 2009) in addition to using some regional floras
(Prescott, 1982). Algal counts were conducted on the most dominant algae found on each
system (cell p/cm2), using the inverted microscope method with sedimentation chambers
(Utermhl, 1931; Lund et al, 1958) and a formula by Wetzel & Likens (1979) modified by
Schwarzbold (1992).

Results

Overall, diatoms dominated the floways in all of the ATS systems studied, in both
biomass and species richness, oscillating from a little over 60% to over 95% of the total biomass
during harvest periods, depending on the system and/or the season. The first two charts on the
following page demonstrate the percent biomass of the different algal groups found on the ATS
systems in Springdale, AR and Gloucester Point, VA, followed by a diagram with species
diversity in April, 2009 of both the 1% slope wood and 2% slope aluminum floways at the
Muddy Run, PA site.

Some of the most abundant and frequent species of diatoms found in the freshwater ATS
systems are Melosira varians, Diatoma vulgare, Fragilaria spp. (ex. F. capucina), Gomphonema
spp. (ex. G. olivaceum and G. truncatum), Navicula spp. (ex. N. gregaria, N. lanceolata), and
Nitzschia spp. (ex. N. cf. dissipata). In the estuarine systems, the diatoms encountered are the
tube-forming Berkeleya rutilans (and some Navicula sp. associated), Melosira nummoloides,
Licmorpha spp., Gyrosigma sp., and Nitzschia sp. Many of these common diatom elements are
shown in the following figures.

As commented previously, diatoms, like other algae, can be found naturally in many
environments, though Gyrosigma sp., Navicula spp., and Nitzschia spp. tend to be common on
the sediment or strictly part of the silt flora, also on rocks which are covered with fine sediment.
Diatoma vulgare and Gomphonema spp. are commonly found growing on rocks (sometimes
plants) in rivers, and are capable of massive growths. Fragilaria spp. are found in various
habitats, like the plankton in lakes or on sediments and plants in rivers. Melosira varians forms
loose flocs on river banks especially in the slower moving parts of the river. These are easily
broken apart, releasing the frustules into the water column of the river, where they are able to
move downstream. M. varians is a freshwater species, however in brackish environments,

52
species such as M. nummoloides can be found in similar habitats (Kelly, 2000). Berkeleya
rutilans grows in estuarine-marine environments, attached to the rocks, forming colonies
resembling macroscopic seaweeds, though having a free living state during times of the year
(Lobban, 1984).

References

Barber, HG & Haworth, EY (1994) A guide to the morphology of the diatom frustules.
Freshwater Biological Association Scientific Publication 44: 1-112.
Bicudo, CEdeM & Menezes, M (org.) (2005) Gneros de Algas de guas Continentais do
Brasil: Chave para Identificao e Descries. So Carlos: RiMa, 489pp.
Biggs, BJF (1996) Patterns in Benthic Algae of Streams. In: Stevenson, RJ; Bothwell, ML;
Lowe, RL eds., Algal Ecology: Freshwater Benthic Ecosystems. San Diego: Academic
Press, p. 31-56.
Cox, EJ (1996) Identification of Freshwater Diatoms from Live Material. London: Chapman &
Hall, 158 pp.
Hoek, C van den; Mann, DG; Jahns, HM (1995) Algae: An introduction to phycology.
Cambridge: Cambridge University, 627pp.
John, DM & Williamson, DB (2009) A Practical Guide to the Desmids of the West of Ireland.
Galway, Ireland: Martin Ryan Institute, 196pp.
Kelly, M (2000) Identification of common benthic diatoms in rivers. Field Studies 9: 583-700.
Komrek, J & Anagnostidis K. (1998) Cyanoprokaryota 1.Teil: Chroococcales. In: Ettl H.,
Grtner, G; Heynig, H; Mollenhauer, D eds, Swasserflora von Mitteleuropa 19/1, Jena:
Gustav Fischer, 548 pp.
Komrek, J & Anagnostidis K. (2005) Cyanoprokaryota 2.Teil: Oscillatorialles. In: Budel, B;
Krienitz, L; Grtner, G; Schagerl, M. eds., Swasserflora von Mitteleuropa 19/2, Mnchen:
Elsevier GmbH, 758 pp.

53
Krammer, K & Lange-Bertalot, H (1986) Bacillariophyceae 1. Teil: Naviculaceae. In: Ettl H.,
Gerloff, J; Heynig, H; Mollenhauer, D eds, Swasserflora von Mitteleuropa 2/1, Stuttgart:
Gustav Fischer Verlag, 876 pp.
Krammer, K & Lange-Bertalot, H (1988) Bacillariophyceae 2. Teil: Bacillariaceae,
Epithemiaceae, Surirelliaceae. In: Ettl H., Gerloff, J; Heynig, H; Mollenhauer, D eds,
Swasserflora von Mitteleuropa 2/2, Stuttgart: Gustav Fischer Verlag, 596 pp.
Krammer, K & Lange-Bertalot, H (1991) Bacillariophyceae 3. Teil: Centrales, Fragilariaceae,
Eunotiaceae. In: Ettl H., Gerloff, J; Heynig, H; Mollenhauer, D eds, Swasserflora von
Mitteleuropa 2/3, Stuttgart/Jena: Gustav Fischer Verlag, 576 pp.
Krammer, K & Lange-Bertalot, H (1991) Bacillariophyceae 4. Teil: Achnanthaceae, Kritische
Ergnzungen zu Navicula (Lineolatae) und Gomphonema Gesamtliteraturverzeichnis Teil 1-
4. In: Ettl H., Gartner, G; Gerloff, J; Heynig, H; Mollenhauer, D eds, Suwasserflora von
Mitteleuropa 2/4, Stuttgart/Jena: Gustav Fischer Verlag, 437pp.
Lobban, CS (1984) Marine tube-dwelling diatoms of eastern Canada: descriptions, checklist,
and illustrated key. Can. J. Bot. 62: 778-794.
Lund, JWG; Kipling, C; Le Cren, ED (1958) The Inverted Microscope Method of Estimating
Algal Numbers and the Statistical Basis of Estimations by Counting. Hydrobiologia 11: 143-
170.
Prescott, GW (1982) Algae of the Western Great Lakes Area. Revised edn. Dubuque, Iowa:
Brown, WMC, 977pp.
Reviers, B de (2006) Biologia e filogenia das algas. Trans. Iara M. Franceschini. Porto Alegre:
Artmed Editora, 280 pp.
Round, FE; Crawford, RM; Mann, DG (1990) The Diatoms: Biology and Morphology of the
Genera. Cambridge: Cambridge University Press, 747pp.
Schwarzbold, A. (1992) Efeitos do regime de inundao do rio Mogi-Guau (SP) sobre a
estrutura, diversidade, produo e estoques do perifton de Eichornia azurea (Sw) Kunth da
Lagoa do Inferno. PhD dissertation, Universidade Federal de So Carlos, So Carlos,
237pp.
Sze, P (1993) A Biology of the Algae. 2nd ed. Dubuque, Iowa: WC Brown Publishers, 259pp.
Utermhl, H (1931) Neue Wege in der quantitativen Erfassung des Planktons. Verh. Internat.
Verein. Limnol. 5: 567-596.
Wetzel, RG & Likens GE (1979) Limnological Analyses. Philadelphia: Saunders, 357pp.

54
 Relative biomass of the different algal groups of the ATS Springdale floway during
the indicated period. Most common species of each group are listed with
percentages.

Relative biomass of the different algal groups of the ATS Springdale floway
during the indicated period. Most common species of each group are listed with
percentages.

55
 Wood and Aluminium Floways species group comparison

70

60

50
Number of Species

40

30

20

10 Wood

Aluminium

0 Total
E

E
IA

E
E

A
EA

EA
EA

EA

EA
EA

YT
R
TE

PH
YC

YC
YC

YC

YC
YC
C

IO

PH
PH

PH

PH
PH
BA

AP

O
O

O
O
O

LA
EM

YD
R

YS

ID
N
N

LO

LE

IL
YA

H
R
N

M
C

AP
G
H

C
G

LA
C

BA
EU
C

ZY

R
H
C

Species richness comparison between the two ATS floways at Muddy Run.

Great Wicomico River ATS - Virginia Great Wicomico River ATS - Virginia
3-D screen 11/4/2009 2-D screen 11/4/2009
Chlorophyt
a
7% Cyanobacte
ria
10%
Chlorophyta
18%

Bacillarioph
Bacillarioph yta
yta 82%
83%

Relative biomass counts conducted on samples from the Great Wicomico River comparing the
2-D and 3-D screens in use. Besides difference in diatom species composition in each sample,
there was a difference of taxonomic groups. More filamentous and erect diatom species found
on the 3-D, compared to more adnate and erect species on the 2D. More samples need to be
analyzed to have a better understanding of the significance of this difference.

56
 1 2

3 4

5 6

1 Berkeleya rutilans; 2 Melosira nummoloides; 3 Nitzschia sp.; 4 Melosira spp.; 5 Cocconeis sp.
on Ceramium sp.; 6 Licmorpha sp.

57
 1 2

3 4

5 6

1 Surirella sp.; 2 Encyonema sp.; 3 Diatoma vulgare; 4 Achnanthes sp.; 5 Eunotia sp. 6 Nitzschia sp.

58
 1 2

3 4

5 6

1-2 Biofilm of diatoms; 3 Cymbella sp..; 4 Gomphonema truncatum; 5 Melosira varians;

6 Pediastrum boryanum

59
 a

1 2 3

4 5 6

b
a

7 8 9
1 Cladophora glomerata; 2 Ulothrix sp.; 3a C. glomerata; 3b Oedogonium sp.; 4 Klebsormidium sp.;
5 Melosira varians; 6 Uronema sp. 7 Phormidium sp.; 8a Spirogyra sp.; 8b Rhizoclonium sp.; 9 Diatoma vulgare

60
 Appendix B

Final Report on the Susquehanna River Algal Turf Scrubber Project

December 2009

Patrick Kangas (1), Walter Mulbry (2), Philip Klavon (1), Dail Laughinghouse (1)

1) Environmental Science and Technology Department

University of Maryland

College Park, Maryland

2) United States Department of Agriculture

Beltsville Agricultural Research CenterBeltsville, Maryland

61
Main Findings of the Report:
Over a growing season of 8 months, net biomass production averaged 14.0 grams dry weight/m2/day
for the aluminum ATS and 11.7 grams dry weight/m2/day for the wooden ATS with a peak of 24.5 grams
dry weight/m2/day on 7/28/08.

These net biomass production values are higher than any published values from natural aquatic plant
communities in the Chesapeake Bay.

Biomass production was composed of the vacuum harvest (52% of the total production for the
aluminum ATS and 29% for the wooden ATS), the greenwater (42% of the total production for the
aluminum ATS and 66% for the wooden ATS) and the slough (6% of the total production for the
aluminum ATS and 5% for the wooden ATS).

During 2009 at peak metabolism the wooden ATS decreased nutrient concentrations to a greater degree
(nitrate-N 27%, orthophosphate-P 50%, total N 37% and total P 17%) than the aluminum ATS (nitrate-N
18%, orthophosphate-P 17%, total N 28% and total P 9%).

Algal biomass averaged between 2-3% nitrogen content and between 0.25-0.35% phosphorus content,
depending on the component (vacuum harvest, greenwater, slough).

Net production based on ecosystem metabolism measurements of the two ATS varied seasonally from
less than 1 to 8.6 grams oxygen/m2/day.

Species diversity of the ATS was high: 195 species from six algal phyla on the aluminum ATS and 196
species from five algal phyla on the wooden ATS

62
NARRATIVE

This is the final report on the Susquehanna River ATS (algal turf scrubber) project. The proposal for
the project was submitted in late spring of 2007 and the intention was to have the ATS operational by
early spring 2008. The summer of 2007 was spent searching for a suitable site with close proximity to
the river, no shading and security. Contact was eventually made with the Exelon Energy Corporation,
which operates several power plants on the lower Susquehanna River. The manager of the Conowingo
hydroelectric plant became interested in the project and she assisted in establishing and conducting the
project since the first contact. Meetings were held with Exelon administrators during the fall and winter
in order to secure corporate permission to operate the ATS on Exelon property. After extensive
communications between the Exelon Corporation, the Smithsonian Institutes National Museum of
Natural History and the University of Maryland, a suitable agreement was established for the research
to be carried out at the Muddy Run hydroelectric power plant in Lancaster County, Pennsylvania. The
Hydromentia Corporation of Ocala, Florida fabricated the ATS from aluminum and had it shipped to the
site in late spring of 2008. The system was constructed at Muddy Run with assistance from Exelon staff
and from the company Living Ecosystems of Easton, Maryland. The dimensions of the original system
were 1 by 300 and it was positioned at a 2% slope.

Construction of the ATS was completed on 6/6/08 with water input from the Muddy Run Reservoir.
Rocks from the Susquehanna River at Muddy Run were added to the ATS to both provide a seed source
for benthic algae and to submerge the ATS screen, which was not physically attached to the trough or
bed of the ATS. Wire braces were also installed in order to keep the screen submerged. Benthic algae
from a local creek, which were placed in bags fashioned from the ATS screen material, was also used to
seed the system. Local streams were searched for Cladophora, which was not found at the Muddy Run
site. Two nearby streams with Cladophora were located and rocks from these streams were added to
the system with the intention of adding this genus to the scrubber community.

An algal community developed relatively rapidly and the system was first harvested on 6/23/08 to
initiate the study of biomass production and nutrient dynamics. Measurements of biomass production,
or net primary production, began on 6/28/08 and continued throughout the growing season
approximately every five-seven days.

An additional ATS was constructed out of wood with fiberglass coating by Living Ecosystems with the
same dimensions as the original aluminum ATS. This system was installed parallel to the aluminum ATS
at a 1% slope in order to investigate the effect of ATS slope on productivity and nutrient dynamics. The
wooden ATS also differed from the aluminum ATS in mesh size of the screen used to grow algae: the
wooden ATS had a mesh size of 0.04 cm2 while the original aluminum ATS had a mesh size of 0.25 cm2.
The different mesh sizes were used to investigate the effort of screen surface area (which was assumed
to be inversely related to mesh size) on productivity. The wooden ATS was constructed in the early fall
2008 and it was first harvested on 10/16/08. Both ATS were operated through the fall and the last
harvests were taken on 12/8/08.

63
 The intention was to operate both ATS through the winter but cold temperatures (see Figure 1)
caused the water in the systems to freeze periodically. Freezing started at the bottom of the ATS and
this caused water to spill on to the ground creating a sheet of ice around the systems. The ice was a
safety hazard so the ATS were shut down several times during the winter and early spring. Thus, no
harvests were taken during the winter because of the freezing problem.

Water flow was restored in March 2009 and algal growth occurred immediately. The systems were
first harvested on 4/14/09. After an annual cycle of data was gathered from the original aluminum ATS
at the end of July 2009, its slope was dropped to 0.5% in order to investigate the effect on productivity
through the late summer and fall. The last harvest of the study occurred on 10/29/09 once an annual
cycle of data had been gathered on the wooden ATS. After this time, emphasis in the project shifted
from field work to analysis of data and to final report preparation.

A four month-long infestation of herbivorous chironomid fly larvae occurred in 2009 that dominated
the ecology of the systems. This infestation was somewhat surprising since no chironomids were
observed in the systems in 2008. The chironomids first appeared in late May 2009 in the aluminum ATS
and they cleared algae from the screens throughout the summer. The larvae were present in the
wooden ATS during this time but they did not cause visible eat-outs until late July. The intensity of the
herbivory was not uniform throughout the ATS but the chironmids did consume enough algae to cause
large areas of bare screen in the systems. Herbivory seemed to be most evident in the lower and middle
sections of the ATS. Several control techniques were used to manage the infestation including altered
harvest frequencies, short term (usually 3 day duration) drainage of the systems and use of a bacterial
insecticide. These control techniques did reduce chironomid densities but the fly larvae always
returned. The chironomids seemed to enter the system through the input water, though it is possible
that egg laying by adults within the raceways could have contributed to the population. The infestation
disappeared in mid-September, presumably due to seasonal completion of the fly life cycle. After this
time algal growth increased and full coverage of the screens developed.

METHODS

Basic Water Quality

Water quality parameters were measured before each harvest in the late morning or afternoon in
order to quantify the maximum effect of metabolism of the algal community. Measurements were
made at the top of the system, where water enters from the Muddy Run reservoir, and at the bottom of
the system, as the water flows into the river. The difference between bottom and top values indicates
the effect of algal metabolism as water passes over the turf. Water temperature, dissolved oxygen
concentration and percent saturation of dissolved oxygen in the water were measured with a YSI meter
and pH was measured with an Accumet meter. On 10 dates, water quality parameters were measured
over a diurnal cycle in order to assess ecosystem metabolism of the turf communities.

64
Biomass Harvest

Algal biomass was harvested in sample sections along the length of the ATS in order to identify
possible longitudinal changes in the community. Several different patterns of harvests were tried
initially. The final pattern of harvest was established on 7/28/08 and it was used throughout the study.
For the aluminum ATS biomass was harvested in two adjacent 10 sections of the scrubber near the top
of the system (T1 at 50 to 60 feet downstream from the surge box, T2 at 60 to 70 feet), in the middle
(M1 at 170 to 180 feet, M2 at 180 to 190) and at the bottom of the system (B1 at 280 to 290 feet and B2
at 290 to 300 feet). This pattern of harvest included replication of sampling (n=2) at three different
locations along the longitudinal gradient of water input at the top of the system to the water output at
the bottom of the system. Each individual section of the aluminum ATS had an area of approximately
0.9 m2 (the one foot wide trough x a 10 section between supports on the trough) while each section of
the wooden ATS had an area of approximately 0.7 m2. For the aluminum ATS the total sample area at
each harvest was then 5.4 m2 which is 20 % of the total scrubber area (27.6 m2) for the aluminum ATS.
At the time of harvest the water input to the ATS was turned off at the inflow faucet and the scrubber
was allowed to drain for between - 1 hour. Harvesting was done with a wet/dry vacuum. Each
sample section of the trough was vacuumed, starting at the top of the system and working sequentially
downstream. The entire ATS area (except for the first section below the surge box which was
maintained as an unharvested refuge) was harvested but only the biomass production samples were
retained for measurements. After a sample section was vacuumed, the resulting slurry was dewatered
by sieving the harvested material through a 3mm mesh nylon netting (Aquatic Ecosystems, Apopka, FL).
The biomass retained in the net was air-dried at 25 degrees C using an electric fan, oven-dried at 70
degrees C for 24 hours and then weighed. Data for biomass production or net primary production were
calculated by dividing the oven-dried mass by the number of days between harvest dates. Throughout
the study 235 vacuum harvest samples were analyzed for the aluminum ATS and 160 were analyzed for
the wooden ATS.

Two other components of the biomass production of the algal turf community were assessed: algal
biomass suspended in the water that was sieved through the nylon netting (termed green water) and
algal biomass that broke off from the turf between harvests (termed slough). For biomass in green
water, the volume of sieved water was measured at the time of harvest and a one liter sample was
collected and returned to the laboratory. The biomass was allowed to settle out of suspension in the
sample bottle creating a dense layer in the bottom of the bottle. The overlying water was removed by
vacuuming or decanting and the remaining slurry that contained the biomass was spread out on a tray
lined with a plastic film in which the water was allowed to evaporate. The biomass remaining after
evaporation was oven-dried at 70 degrees C and weighed. Biomass production of the green water
component was calculated by dividing the biomass by the area of turf that was harvested and sieved
and by the number of days between harvest dates. Routinely, green water was combined for paired
samples at the top, middle and bottom of the ATS resulting in three samples for each harvest date after
7/28/08. Throughout the study 105 greenwater samples were analyzed for the aluminum ATS and 68
were analyzed for the wooden ATS. Biomass that sloughed from the turf between harvest dates was

65
collected in a nylon mesh bag that was attached to the outlet drain pipe at the bottom of the ATS.
Material collected in the bag was removed at the time of harvest and it was processed by the same
procedure as the biomass harvest collected with the wet/dry vacuum. To calculate the slough
production, the biomass collected between harvest dates was divided by the total area of the scrubber
and by the number of days between harvests. Throughout the study 23 slough samples were analyzed
for the aluminum ATS and 10 were analyzed for the wooden ATS.

Ecosystem Metabolism

Diurnal curve analysis of dissolved oxygen concentration was used to estimate ecosystem metabolism
and to help assess the functioning of the ATS systems. Nine diurnal curves were analyzed during the
study period covering all of the seasons of the year. The basic water quality parameters of temperature,
dissolved oxygen concentration, percent saturation of oxygen and pH were measured at various time
intervals over a 24-hour time period to create the diurnal curves. Ecosystem metabolism was estimated
using the dissolved oxygen concentration data. The concentration at the top of the system (input water)
was subtracted from the concentration at the bottom of the system (output water) and this difference
was divided by the turnover time of water in the system (an average of 10 minutes was used) to
calculate rates of oxygen change. These rates of change in oxygen concentration were then plotted over
24 hours and graphically integrated to estimate metabolism. This approach is essentially a modified
version of the standard upstream-downstream technique for measuring ecosystem metabolism in
flowing water systems (Odum 1956).

No corrections to the rate-of-change curves were made for ambient diffusion based on three reasons.
First, percent saturation of dissolved oxygen was usually near 100% when top (input) and bottom
(output) samples were averaged. Second, it was assumed that the fast turnover time of water in the
system (about 6-20 minutes) precludes any major changes in oxygen concentration due to diffusion.
Third, upstream-downstream sampling of dead turfs, either during the winter or after drainage for
chironomid control in the summer of 2009, indicated little evidence of physical diffusion.

Nutrients

Water samples were collected at the inflow and at the outflow of the ATS before harvests. These
samples were acidified and stored, usually at 4 degrees C, prior to analysis for total nitrogen (TN),
nitrate-nitrogen (NO3-N), total phosphorous (TP) and orthophosphate-phosphorous (PO4-P). The water
samples were also analyzed ammonium-nitrogen but the resulting values were below detection limits or
non-interpretable. Biomass samples were analyzed for TN and TP after oven-drying.

Algal Community Structure

Samples of algae from the scrubber were collected periodically and examined with a compound
microscope for the purpose of describing the structure of the community at the time of almost every
harvest. Two or three samples from each turf along with a sample of the green water and a sample of
the slough were routinely collected during each harvest and these were examined to rapidly assess the

66
dominant alga taxa and the overall condition of the system. Samples were collected for a more in-depth
assessment of the species composition of the turfs on 7/23/08, 8/1/08, 10/27/08, 4/7/09, 6/12/09,
7/17/09, 10/9/09. These samples were carefully examined in order to construct a species list of algal
taxa found on the scrubber.

RESULTS

Water Quality and Quantity

Water input to both of the ATS came from the Muddy Run Reservoir, which is filled from and drained
to the Susquehanna River on a daily basis in the operation of Exelons pumped-storage hydroelectric
facility. The quality of the input water to the ATS was thus modified from the ambient river conditions
by a relatively short retention time in the reservoir. At one seasonal extreme, during the summer, input
dissolved oxygen concentration was low at around 4 mg/l with about 50% saturation and pH was about
7.5. At the other seasonal extreme, during the winter, input dissolved oxygen concentration was high at
around 12 mg/l with about 90% saturation and pH was about 8.0. Passage of the input water through
the ATS increases dissolved oxygen concentration, percent oxygen saturation and pH in proportion to
ecosystem metabolism of the ATS, as will be discussed in a later section of this report. During mid-day
these increases are dramatic with dissolved oxygen concentration and percent saturation more than
doubling and pH increasing by nearly two units from top to bottom of the ATS raceways. These
increases are especially noteworthy given that the turnover time of water in the systems was between 7
and 10 minutes.

Average nutrient concentrations (n=17) of the input water were as follows: 0.89 mg/l NO3-N, 1.38
mg/l TN, 0.06 PO4-P, and 0.12 mg/l TP. These nutrients were reduced by uptake in algal growth as the
water passed along the ATS raceway. Mid-day removal percentages are given in Table 1. These data are
point measurements within the diurnal cycle and they represent the maximal uptake rates since uptake
is proportional to ecosystem metabolism and ecosystem metabolism is highest at mid-day.

Data on water flow rates and turnover times for the raceways are given in Table 2. The flow rate was
about 1 liter/second throughout the study but it was reduced slightly after the wooden ATS was
constructed in late 2008. The water turnover rates were relatively fast at between 6 and 20 minutes.
This represents the time it takes water to move from the top to the bottom of the raceway on average.
Turnover time was strongly influenced by water depth, which was inversely related to the slope of the
raceway.

Biomass Production

There were three components of algal biomass production for the ATS based on the methods
described above: the biomass harvested with the wet/dry vacuum (vacuum harvest), the biomass
contained in the water drained from the vacuum harvest (green water) and the biomass that sloughed
off the turf between harvests and that was collected in the net at the bottom of the system (slough).
Overall biomass production data are summarized in Tables 3a and 3b with seasonal and growing season

67
averages. Average total daily production was slightly higher for the aluminum ATS (14.0 g/m2/day) in
comparison with the wooden ATS (11.7 g/m2/day). Peak production occurred in the summer months
for both ATS. The highest daily production occurred on the aluminum ATS on July 28, 2008 at 24.5
g/m2/day.

Vacuum harvest contributed about half of the growing season production for the aluminum ATS and
about 30% of the total for the wooden ATS. These samples were dominated by a mixture of filamentous
green algae (Spirogyra was dominant) and filamentous diatoms (Melosira was dominant) along with
pennate diatoms and bluegreen algae (Oscillatoria and Phormidium).

Green water biomass production contributed 42% of the aluminum ATS production and 66% of the
wooden ATS production. These samples were strongly dominated by fragments of filamentous diatoms
and by pennate diatoms, with practically no filamentous green algae. Averaged over the entire study,
average greenwater concentration was 14.2 g dry weight/liter and average greenwater volume was 7.4
liters/m2 of turf for the aluminum ATS. Average greenwater concentration was 11.3 g dry weight/liter
and average greenwater volume was 8.0 liters/m2 of turf for the wooden ATS.

Slough biomass production contributed about 5% of the total production for both ATS. These
samples were strongly dominated by the filamentous green alga, Spirogyra, though all species from the
turf were present in small amounts. This portion of the overall production was the most difficult to
measure and more study is required for adequate quantification. Problems occurred when the slough
bag would become detached between harvests. Also, in the fall season deciduous leaves from nearby
trees fell and collected in the slough bags and it was not possible to physically separate leaves from the
algae in order to quantify slough.

Nutrient Uptake Rates

Nutrient uptake by the ATS was found by multiplying biomass production rate (g dry
weight/m2/day) by the nutrient contents of the biomass (%). These data are illustrated for the
aluminum ATS in Figures 2 and 3. Data for biomass production in these figures is from Table 3a.
Although the magnitudes of nutrient contents differed between the different components of biomass
production, the ratios between nitrogen (N) and phosphorous (P) are about 8 to 1 for the different
components. For the aluminum ATS total uptake rates are 0.34 g N/m2/day and 0.04 g P/m2/day.

Ecosystem Metabolism

Diurnal curve analysis of ecosystem metabolism provides a holistic view of the functioning of the ATS.
Dissolved oxygen concentration at the bottom of the systems rises during the day due to photosynthesis
and falls at night due to respiration. pH at the bottom of the systems also has this diurnal pattern of
change but this reflects dissolved CO2 dynamics since pH is inversely proportional, in a non-linear
fashion, to CO2 concentration dissolved in the water.

68
 Values for net production from the diurnal curves are given in Table 4. Values range from about 1
gram of O2/m2/day during the winter months to nearly 10 grams of O2/m2/day in the summer months.
An interesting feature of most of the diurnal curves is the minor role played by community respiration in
the ATS. Respiration occurs when oxygen change is negative. Day-time respiration is masked by net
production, but at night the rate-of-change of oxygen should become negative. Although some negative
rates-of-change did occur on some of the diurnal curves, these rates were always relatively low. In part,
the negative rates-of-change at night were mitigated by apparent turbulent diffusion at the top of the
system due to the wave generation mechanism. This effect was measured to be an immediate increase
of 1 2 mg/l of oxygen once the input water enters the system. Corrections for this increase were made
in the analysis of the curves. Thus, total oxygen increase at the bottom of the system is partly due to
ecosystem metabolism and partly due to physical, turbulent diffusion.

Algal Community Structure

A total of 195 species were identified from six algal phyla on the aluminum ATS over the study period
(Figure 4) and a total of 196 species were identified from five algal phyla on the wooden ATS (Figure 5).
Many of the species were rare in the context of frequency of occurrence: 103 species or 53% of the total
community were found only once (a frequency of 1/26 or 4%) on the aluminum ATS and 92 species or
47% of the total community were found only once (a frequency of 1/22 or 5%) on the wooden ATS.
Thus, as is typical of any ecological community, most of the species in the algal turf community were
rare.

Five species were found in at least half of the samples on the aluminum ATS (from the genera
Navicula, Cyclotella, Spirogyra, Ulnaria and Melosira) and nine species were found in at least half of the
samples on the wooden ATS (from the genera Navicula, Diatoma, Ulnaria, Nitzschia, Cyclotella, Melosira,
Spirogyra, Syndra and Monorapidium).

DISCUSSION

Overall Productivity

The two ATS raceways at Muddy Run had net biomass productivities that were comparable to other
published outdoor studies from the Chesapeake (Mulbry et al. 2008, in press). Growing season averages
were also higher than any reported net productivity from a natural aquatic plant community from the
Chesapeake Bay (Table 5). The amplified productivities of the ATS over the natural communities is
presumably due to the energy subsidies of the engineered design that the ATS receive from the artificial
substrate and pulsing water flows but high species diversity may also play a role as will be discussed in
the next section.

However, there are reasons to question whether the ATS productivities measured at Muddy Run
might be less than what could have been achieved. General predictions for ATS productivity suggested
that >30 grams dry wt./m2/day might be possible (Adey and Loveland 2007). In fact, higher
productivities have been measured in similar ATS raceways in the more southerly portions of the bay on

69
the Wicomico River in Virginia (W. Adey, personal communication) and at the Virginia Institute of
Marine Science at the mouth of the York River (E. Duffy, personal communication).

One possible cause of reduced productivity in the Muddy Run ATS studied here was artificial shading
from the walls of the raceways. A preliminary study was made in the late fall 2008 to test this
hypothesis. Shading by the walls may be suspected to be significant at this season, due to the reduced
angle of the sun in the sky. Algal biomass was harvested from five sections of the aluminum ATS and
then these sections were allowed to grow back. At the next harvest biomass in the shaded half of the
raceway was gathered separately from biomass in the unshaded half of the raceway. After processing,
the results indicated 20% greater production in the lighted half of the system compare to the shaded
half. Thus, shading may well be limiting productivity of the ATS, at least during certain times of the year.

One remarkable finding of this study was the relative lack of effect of chironomid herbivory on overall
productivity of the ATS. This observation comes from comparing productivity during the summer of
2008, when no chironomids were found on the ATS, to productivity during the summer of 2009, when
chironomids were abundant and their feeding created bare areas with no filamentous algae on the
screens. Apparently, productivity shifted from filamentous species to single-celled species during the
chironomid outbreak and this pattern is indicated to some extent by the changes in proportion of the
vacuum harvest (which was dominated by filamentous species) in relation to the greenwater harvest
(which may have had greater contribution from single-celled pennate diatoms). This pattern of biomass
harvest productivity and herbivory was also found with oxygen-based productivity estimates. In
general, there seemed to be some indication that the greenwater proportion of total productivity
increases with stress, as noted above with chironomid herbivory, since this proportion was higher in the
spring and fall when temperature and/or light levels may be limiting the metabolism of the system. The
proportions of total productivity from vacuum harvest, greenwater and slough will have potentially
significant implications for commercial-scale harvesting techniques, so this topic deserves further study.

Another hypothesis about limitations on Muddy Run ATS productivities concerned the slough
component of biomass harvest. One idea was that slough biomass was being lost in the outlet water
and that there was biomass production missing from the totals reported in Table 3. Slough was difficult
to measure due to problems with the bags attached to the outlet pipe at the bottom of the raceways.
To test the hypothesis of missing slough biomass production, short-term (5 minutes) collections of
slough were made over four diurnal cycles in summer-fall 2009. These data averaged 1.4 grams dry
wt./m2/day for the aluminum ATS and 1.0 grams dry wt./m2/day for the wooden ATS. These values are
consistent with the longer term (5-7 days) collections of slough reported in Tables 3a and 3b so the
hypothesis is not supported. The only other possibility is that the missing slough passed through the
collection bags, both during the short-term and the long-term collections. If this situation is occurring,
finer mesh collection bags will be required to sample this missing biomass and alternative processing
techniques will be required to collect, dry and weigh it.

70
Algal Diversity

The sheer magnitude of the diversity of algal species found on the two ATS at Muddy Run is amazing.
Nearly 300 species were found on the two scrubbers, which together make up only about 60 m2 of
surface area. To some extent the high diversity of algal species on the ATS is due to the species diversity
in the source communities for the systems. Ultimately, the source of species is the Susquehanna River,
which is a lotic or flowing water ecosystem. However, the river water is passed through the Muddy Run
Reservoir before it reaches the ATS. The reservoir is a lentic or standing water ecosystem. Thus, since
different species are adapted to lotic vs. lentic ecosystems, these two sources probably contribute to the
high diversity of the scrubbers.

Another factor that may have contributed to the species diversity on the scrubbers is the spatial
heterogeneity of the systems. The scale of the systems is actually large, due to their 300 length, and
water flowing over this scale dimension created a mixture of regular, longitudinal zones and irregular,
scattered patches. This spatial heterogeneity offers more opportunities for species colonization than
would an equivalent surface area of a uniformly distributed environment.

The vertical structure of the turf community may contribute to the species diversity of the systems.
The character of the algal communities is dominated by filamentous species, more-or-less attached to
the screen, that form a canopy over the turf. Most of the rest of the community are single-cell
understory species, some epiphytic on the filaments but many others just growing unattached among
the filaments. Thus, the multi-celled filaments themselves contribute to the diversity but they also may
facilitate the presence of single-celled species.

Finally, there is a heterogeneity in time imposed on the scrubbers due to seasonal change in
temperature and other factors. As with other ecological communities, a seasonal succession of species
occur since different species are adapted to the different conditions of the seasons. Since the sampling
of the turf communities spanned at least one annual cycle, the seasonal succession of species
contributed to the diversity of the systems.

Understanding the causes of the high species diversity of the Muddy Run ATS deserves more study.
The diversity seems to be higher than would be expected from any single natural algal community. A
further question is whether or not the high diversity contributes to the high productivity of the ATS
relative to natural communities as was discussed earlier. This kind of question illustrates the
importance of studying the ecology of these systems in order to maximize their functioning in nutrient
removal and biofuel feedstock production.

71
ACKNOWLEDGEMENTS

Mary Helen Marsh, past Manager of Exelon Corporations hydroelectric facilities on the Susquehanna
River was a catalyst from the start to the finish of the ATS project. Thomas Jenkins of Exelon provided
invaluable assistance with the construction and operation of the ATS at Muddy Run. Tim Goertemiller,
of Living Ecosystems in Easton, Maryland, helped with set up of the aluminum ATS and his firm designed
and contructed the wooden ATS.

LITERATURE CITED

Adey, W. and K. Loveland. 2007. Dynamic Aquaria: Building Living Ecosystems, 3rd edition. Elsevier,
Amsterdam.

Boynton, W. R., W. M. Kemp and C. W. Keefe. 1982. A comparative analysis of nutrients and other
factors influencing estuarine phytoplankton production. Pp. 69-90. in: Estuarine Comparisons. V. S.
Kennedy (ed.). Academic Press, New York, NY.

Cahoon, D. R. 1975. Net Productivity of Emergent Vegetation at Horn Point Salt Marsh. M. S. Thesis,
University of Maryland, College Park, MD.

Harding, L. W., Jr., D. Degobbis and R. Precali. 1999. Production and fate of phytoplankton: Annual cycles
and interannual variability. Pp. 131-172. in: Ecosystems at the Land-Sea Margin: Drainage Basin to
Coastal Sea. American Geophyical Union.

Johnson, M. 1970. Preliminary Report on Species Composition, Chemical Composition, Biomass and
Production of Marsh Vegetation in the Upper Patuxent Estuary, Maryland. Chesapeake Biological
Laboratory, Solomons, MD. Reference No. 70-130.

Mulbry, W., S. Kondrad, C. Pizarro and E. Kebede-Westhead. 2008. Treatment of dairy manure effluent
using freshwater algae: algal productivity and recovery of manure nutrients using pilot-scale algal turf
scrubbers. Bioresource Technology 99:8137-8142.

Mulbry, W., P. Kangas and S. Kondrad. In press. Toward scrubbing the Bay: Nutrient removal using small
algal turf scrubbers on Chesapeake Bay tributaries. Accepted for publication in Ecological Engineering.

Odum, H. T. 1956. Primary production in flowing waters. Limnology and Oceanography 1:102-117.

Sellner, K. G. and M. E. Kachur. 1987. Phytoplankton. Pp. 12-37. in: Ecological Studies in the Middle
Reach of Chesapeake Bay. Springer-Verlag, Berlin.

72
Stevenson, J. C. 1988. Comparative ecology of submersed grass beds in freshwater, estuarine, and
marine environments. Limnology and Oceanography 33:867-893.

Turner, R. E. 1976. Geographic variations in salt marsh macrophyte production: a review. Contributions
in Marine Science 20:47-68.

Wass, M. and T. Wright. 1969. Coastal Wetlands of Virginia. Interim Report to the Governor and General
Assembly. Special Report in applied Marine Science and Oceanographic Engineering No. 10. Virginia
Institute of Marine Science, Gloucester Point, VA.

Table 1. Average nutrient removal efficiencies measured at mid-day during spring-fall 2009 (n=17).
These data were calculated by dividing the output concentrations (at the bottom of the systems) by the
input concentrations (at the top of the systems).

Nutrient component Aluminum ATS Wood ATS

N-NO3 18% 27%

P-PO4 17% 50%

Total N 28% 37%

Total P 9% 17%

73
Table 2. Water quantity data from the raceways at Muddy Run. Turnover time was calculated based on
a raceway volume calculated by multiplying raceway length (9144 cm or 300 feet) by raceway width (30
cm or 1 foot) by water depth.

Raceway flow rate, average water water turnover

liters/sec depth, cm time, minutes

Aluminum ATS

at 2% slope in

2008 with full

flow 1.10 1.5 6.2

Aluminum ATS

at 2% slope in

2009 with split

flow 0.83 1.5 8.3

Wood ATS in

2009 at 1% slope

with split flow 0.73 2.9 18.2

Aluminum ATS

at 0.5% slope in

2009 with split

flow 0.83 3.7 20.4

74
Table 3a. Biomass production of the Aluminum ATS at 2% slope. Data are grams dry weight/m2/day.
Numbers in parentheses are the months included in the seasons.

Season vacuum harvest greenwater slough total

Summer 2008 14.2 2.4 1.4 18.0

(6-8)

Fall 2008 6.2 4.2 0.5 10.9

(9-11)

Winter 2008-2009 --- --- --- ---

(12-3)

Spring 2009 4.0 8.6 0.7 13.3

(4-5)

Summer 2009 11.6 4.6 1.4 17.6

(6-7)

Fall 2009* 6.6 6.3 0.7 13.6

(8-10)

Growing Season

Averages 7.3 5.8 0.9 14.0

(fall 08-summer 09) (52%) (42%) (6%)

* at 0.5% slope

75
Table 3b. Biomass production of the Wooden ATS. Data are grams dry weight/m2/day. Numbers in
parentheses are the months included in the seasons.

Season vacuum harvest greenwater slough total

Fall 2008 1.8 6.2 ---

(10-11)

Winter 2008-2009 --- --- ---

(12-3)

Spring 2009 1.3 10.3 ---

(4-5)

Summer 2009 4.8 6.6 0.9 12.3

(6-8)

Fall 2009 4.2 6.1 0.2 10.5

(9-11)

Growing Season

Averages 3.4 7.7 0.6 11.7

(spring 09-fall 09) (29%) (66%) (5%)

76
Table 4. Net primary production for the Muddy Run ATS raceways as measured by diurnal oxygen curve
analysis. Data are in units of grams O2/m2/day.

Dates aluminum ATS wooden ATS

9/17/08-9/18/08 7.2 ---

10/29/08-10/30/08 2.5 3.2

12/18/08-12/19/08 1.5 0.8

4/7/09-4/8/09 1.1 1.0

6/11/09-6/12/09 2.4 3.8

7/30/09-7/31/09 8.6 6.8

8/13/09-8/14/09 6.5 6.4

9/25/09-9/26/09 8.2 7.4

11/12/09-11/13/09 3.0 2.3

77
Table 5. Comparison of growing season productivities for the ATS systems at Muddy Run and various
natural aquatic plant communities of the Chesapeake Bay. Data for natural communities have been
converted from various units to grams dry weight/m2/day, according to the footnotes given at the
bottom of the table.

Plant community type Net productivity, Reference

g dry weight/m2/day

Aluminum ATS at Muddy Run 14.0 this study

Wooden ATS at Muddy Run 11.7 this study

Submerged aquatic vegetation* 2 8 Stevenson 1988

Emergent marshes** 11.3 Wass and Wright 1969

8.8 Johnson 1

4.9 Cahoon 1975

3.1 Turner 1976

Phytoplankton*** 2.5 Selner and Kachur 1987

2.0 Boynton et al. 1982

1.5 Harding et al. 1999

* calculated assuming dry weight is 50% carbon.

** calculated assuming that annual belowground net production equals annual aboveground net
production and assuming a 9 month growing season.

*** calculated assuming dry weight is 50% carbon.

78
Figure 1. Water temperature recorded at the top and bottom of the aluminum ATS over the study period.

79
Figure 2. Summary of production and nitrogen uptake on the aluminum
ATS for 2009
80
Figure 3. Summary of production and phosphorus uptake on the aluminum
ATS for 2009
81
Figure 4. Cumulative taxonomic species richness of the aluminum ATS over the
study period.

82
 Appendix C
Ozark Highland Algal Turf Scrubber:
Project Report

December 4, 2009

Prepared by:
The Center for Agricultural and Rural Sustainability
University of Arkansas Division of Agriculture
203 Engineering Hall
Fayetteville, AR 72701

83
 TABLE OF CONTENTS

I. Introduction ..........................................................................................................1

II. Monitoring Strategy .............................................................................................1

a. Water Quality Analysis

b. Biomass analysis

c. Weather Monitoring

III. Data and Results ..................................................................................................2

a. Dry Weight Data ............................................................................................2

b. Water Quality Analysis ..................................................................................4

c. Weather Data .................................................................................................10

IV. Analysis................................................................................................................11

V. Conclusions ..........................................................................................................14

VI. Appendix A: Anion Data

a. Orthophosphate ..............................................................................................i

b. Nitrate-nitrite..................................................................................................iii

VII. Appendix B: Pictures ...........................................................................................xxi

84
 LIST OF FIGURES AND TABLES
Table 1: ATS Biomass Productivity Rates by Section and Averaged by Date ...............3

Table 2: Average Temperature Data by Deployment for the ATS System .....................5

Table 3: Average Dissolved Oxygen Data by Deployment for the ATS System ............6

Table 4: Average pH Data by Deployment for the ATS System .....................................6

Table 5: Average Specific Conductivity Data by Deployment for the ATS System.......7

Table 6: Total Nitrogen Concentrations Above and Below the ATS System, with Percent

Removal ...........................................................................................................................7

Table 7: Total Phosphorus Concentrations Above and Below the ATS System, with Percent

Removal ...........................................................................................................................8

Table 1A: Orthophosphate Concentrations Above and Below the ATS System, with Percent

Removal ...........................................................................................................................i

Table 2A: Nitrate-Nitrite Concentrations Above and Below the ATS System, with Percent

Removal ...........................................................................................................................iii

Figure 1: Average Biomass Production Rates for Each Harvest Period. The line is the seven

day moving average, and the bars represent upper and lower 90 percent confidence intervals.

..........................................................................................................................................2

85
Figure 2: ATS System Growth Profile from the Upper to Lower Section for Harvest Date

August 24, 2009 ...............................................................................................................3

Figure 3: Dissolved Oxygen Data: August 19, 2009 August 24, 2009.........................4

Figure 4: pH Data: August 19, 2009 August 24, 2009 .................................................4

Figure 5: Total Phosphorus Percent Removal from the ATS System .............................8

Figure 6: Total Nitrogen Percent Removal from the ATS System ..................................9

Figure 7: Daily Orthophosphate Removal Across Two Harvest Periods at the ATS ......9

Figure 8: 2009 Precipitation Data Collected Daily for Springdale, AR ..........................10

Figure 9: 2009 Maximum and Minimum Ambient Temperature Data Collected Daily for

Springdale, AR .................................................................................................................10

Figure 10: Biomass Production and Orthophosphate Removal Over the 9-month Harvest

Period at the ATS .............................................................................................................11

Figure 11: Biomass Production and Nitrate Removal Over the 9-month Harvest Period at the

ATS ..................................................................................................................................11

Figure 12: Biomass Production and Total Phosphorus Removal Over the 9-month Harvest

Period at the ATS .............................................................................................................12

Figure 13: Biomass Production and Total Nitrogen Removal Over the 9-month Harvest

Period at the ATS .............................................................................................................12

86
Figure 14: Biomass Production and Maximum Ambient Temperature Over the 9-month

Harvest Period at the ATS ...............................................................................................13

Figure 15 Biomass Production and Average Precipitation Over the 9-month Harvest Period at

the ATS ............................................................................................................................13

Figure 1A: Orthophosphate Removal Over the 9-month Harvest Period at the ATS .....ii

Figure 2A: Nitrate-Nitrite Removal Over the 9-month Harvest Period at the ATS ........iv

87
 INTRODUCTION
The University of Arkansas Center for Agricultural and Rural Sustainability (CARS) designed
and constructed an algal turf scrubber (ATS) in the Ozark Highland Region as part of a project
scope of work with the Smithsonian Institute. CARS operated and monitored the system for one
year. Construction began in May of 2008, and was completed in November of the same year.
However, due to a combination of inclement weather and equipment and power failures, the
system was not operating continuously until March 1, 2009. This report summarizes the biomass
production, water quality, and weather data collected during the nine month operating time.

MONITORING STRATEGY
The sampling regime consisted of weekly monitoring of the system biomass production and
water quality. The sample regime included water quality analysis, biomass analysis, and weather
monitoring.

Water Quality Analysis

A set of multiple parameter water quality monitoring devices (data sondes) were used to monitor
pH, dissolved oxygen, specific conductivity, and water temperature. The data sondes took
measurements every 10 minutes during the deployment period. In addition to the continuous
monitoring of these parameters, water samples were taken every harvest period from the inside
of the surger at the intake, and from the sampling reservoir at the outflow. The water samples
were frozen and then analyzed for orthophosphate and nitrate-nitrite using ion chromatography,
as well as total phosphorus and total nitrogen at the University of Arkansas. In addition to the
weekly sampling events, a set of samples were taken every day for 14 days during the month of
July to establish the changes in water quality and nutrient removal across harvest periods.

Biomass Analysis
The algal biomass was harvested every week using a standard wet/dry shop vac. The algae was
harvested from three 10-foot sections at the upper, middle, and lower sections of the flow-way,
and then sent to the Ecological Engineering Laboratory at the University of Arkansas for dry
weight analysis. The dry weight was determined by drying each sample at 105C for 24 hours.
From this, the dry weight per unit of area produced by the system was determined. The slough
coming off the system was also collected throughout the harvest period, and the dry weight was
determined along with the algae samples taken from the harvest. In addition, a ten centimeter
square section of mesh was removed periodically for taxonomic analysis.

Weather Monitoring
Weather data was collected through local weather monitoring stations. Precipitation, maximum
and minimum temperatures, average humidity, and dew point were recorded for each day.

88
 DATA AND RESULTS
DRY WEIGHT ANALYSIS

Algal biomass was collected from three sections on the flow-way from March 2009 through August
2009. The productivity rates determined for each section (and adjusted for the slough yield per unit area)
are listed in Table 1. The average productivity over time is shown in Figure 1. In addition to repeated
sampling at the upper, middle, and lower sections, the amount of variation in growth across the entire
length of the flow-way was also measured. On August 24, 2009 the system was harvested at alternating
10-foot intervals down the flow-way. A plot of the system profile is shown in Figure 2. It is important
to note that three major system shutdowns occurred during the biomass sampling period. The first
shutdown was in early May, and was the result of theft/vandalism. The second shutdown occurred at the
end of May, and was the result of a power failure associated with the incident that occurred earlier in the
month. The third occurred in early October, and was the result of a power failure at the wastewater
treatment plant. This situation also coincided with heavy rains, and resulted in a lack of data for the first
half of October. In addition to the system shutdowns, there were two instances of irregular harvest
procedure, the first on July 2nd, and the other on August 4th. On these days, thunderstorms interrupted the
harvest session, and the harvest was halted after collecting the biomass samples for dry weight analysis.
The following day, the entire flow-way was vacuumed to ensure that there was no irregularity in the
growing time across the trough.

90.00
Average Production Rates (dw g/sq m/day)

80.00

70.00

60.00

50.00

40.00

30.00

20.00

10.00

0.00
3/1/2009 4/20/2009 6/9/2009 7/29/2009 9/17/2009 11/6/2009

Figure 1: Average Biomass Production Rates for Each Harvest Period. The line is the seven day moving average, and the bars
represent upper and lower 90 percent confidence intervals.

89
 Production (g/sq m/day)

Date # Days Upper Middle Lower Average

3/10/2009 10.00 9.99 3.60 6.80

3/17/2009 7.00 9.32 9.52 9.97 9.60

3/30/2009 13.00 22.45 21.70 21.26 21.81

4/11/2009 12.00 21.35 9.72 18.98 16.69

4/16/2009 5.00 40.40 17.66 31.66 29.91

4/23/2009 7.00 4.69 5.96 15.98 8.88

5/23/2009 7.00 11.83 8.70 16.14 12.22

6/9/2009 9.00 17.41 13.52 22.75 17.89

6/18/2009 9.00 37.17 14.26 55.99 35.81

6/27/2009 9.00 36.78 30.30 36.08 34.39

7/2/2009 5.00 94.56 17.88 42.88 51.77

7/7/2009 5.00 35.22 16.08 15.47 22.25

7/11/2009 4.00 40.79 22.75 23.13 28.89

7/15/2009 4.00 46.19 14.57 18.92 26.56

7/20/2009 5.00 41.63 38.55 60.67 46.95

7/27/2009 7.00 36.44 34.51 19.63 30.19

21
8/4/2009 6.00 50.42 37.48 34.01 40.64
19
Algal Density ( g dw/ft2 )

17 8/10/2009 6.00 16.72 9.56 15.20 13.83

15
8/19/009 9.00 13.50 11.67 19.07 14.74
13
8/24/2009 5.00 40.93 31.31 26.64 32.96
11
9 8/31/2009 6.00 61.76 31.53 63.66 52.31
7
9/9/2009 9.00 39.58 17.78 20.39 25.91
5
0 9/14/2009 50 5.00 10035.95 150
24.44 200
15.60 250
25.33 300
Location on Floway (ft)
9/29/2009 5.00 32.00 32.65 11.86 25.50

Figure10/21/2009
2: ATS System Growth
5.00 Profile from the Upper to Lower
12.66 35.39 Section for Harvest Date 31.08
45.19 August 24, 2009

10/29/2009 8.00 19.78 12.20 17.01 16.33

90
11/4/2009 6.00 9.09 14.85 8.86 10.93

11/9/2009 5.00 16.60 19.72 18.34 18.22

Table 1: ATS Biomass Productivity Rates by Section and Averaged by Date

WATER QUALITY ANALYSIS

The dissolved oxygen, pH, temperature, and conductivity levels were monitored daily at the system
inflow and outflow. A total of 14 successful one-week data sonde deployments were performed from
March through November.

Figures 3 and 4 show pH and dissolved oxygen percent saturation data recorded during the harvest period
from August 19, 2009 through August 24, 2009. The trends shown in these graphs are typical of pH and
dissolved oxygen levels measured during a harvest period.

300
250
DO Saturation (%)

200
150
100
50
0
8/19/2009 8/20/2009 8/21/2009 8/22/2009 8/23/2009 8/24/2009 8/25/2009
Outflow Inflow

Figure 3: ATS System Dissolved Oxygen Data: August 19, 2009 August 24, 2009

10.5
10
9.5
9
pH

8.5
8
7.5
7
8/19/2009 8/20/2009 8/21/2009 8/22/2009 8/23/2009 8/24/2009 8/25/2009
Outflow Inflow

Figure 4: ATS System pH Data: August 19, 2009 August 24, 2009
91
For each parameter the maximum and minimum values, along with the averages and standard deviations,
from each data sonde deployment are given in Tables 2 through 5.

Temperature
Upper Lower Summary
Deployment # Deployed Retrieved Max Min S.D. Max Min S.D. xmin xmax
1 3/9/2009 3/12/2009 24.63 13.53 15.89 1.12 19.63 12.91 15.51 1.67 5.00 0.62
2 3/31/2009 4/6/2009 17.38 11.54 14.00 1.31 20.38 8.55 12.95 2.75 3.00 2.99
3 4/14/2009 4/20/2009 17.39 13.28 15.29 1.00 24.05 11.72 15.45 2.08 6.66 1.56
4 4/21/2009 4/28/2009 20.54 14.88 17.99 1.14 29.05 14.16 18.81 2.68 8.51 0.72
5 5/18/2009 5/26/2009 21.92 16.26 19.25 1.31 26.77 15.58 20.02 2.22 4.85 0.68
6 6/4/2009 6/8/2009 22.95 18.91 20.77 1.10 25.69 18.34 21.28 1.90 2.74 0.57
7 6/9/2009 6/15/2009 32.77 19.79 22.16 1.30 31.79 17.98 23.13 2.73 0.98 1.81
8 7/14/2009 7/20/2009 28.10 21.34 24.88 1.54 32.80 20.55 25.84 2.99 4.70 0.79
9 7/23/2009 7/27/2009 26.46 22.14 24.01 1.19 30.18 21.36 24.79 2.28 3.72 0.78
10 8/7/2009 8/11/2009 30.45 21.28 24.51 2.04 30.95 21.62 25.11 2.67 0.50 0.34
11 8/19/2009 8/24/2009 26.74 20.53 23.73 1.43 31.22 19.89 24.21 2.33 4.48 0.64
12 9/3/2009 9/9/2009 19.32 14.79 17.69 1.01 21.66 12.50 16.76 1.85 2.34 2.29
13 10/7/2009 10/14/2009 21.30 11.60 16.98 2.42 22.68 5.55 15.40 3.96 1.38 6.05
14 11/6/2009 11/20/2009 19.72 14.65 17.40 1.20 21.11 12.79 16.55 1.82 1.39 1.86

Table 2: Average Temperature Data by Deployment for the ATS System

92
 DO Percent Saturation
Upper Lower Summary
Deployment # Deployed Retrieved Max Min S.D. Max Min S.D. xmax xmin
1 3/9/2009 3/12/2009 113.00 32.80 94.35 9.88 205.20 85.20 110.85 27.79 92.20 52.40
2 3/31/2009 4/6/2009 110.60 79.60 94.65 6.66 194.40 54.50 115.93 34.92 83.80 25.10
3 4/14/2009 4/20/2009 112.80 51.20 94.30 7.73 159.50 69.60 104.66 20.70 46.70 18.40
4 4/21/2009 4/28/2009 118.80 63.70 94.97 11.14 201.60 47.20 118.78 38.87 82.80 16.50
5 5/18/2009 5/26/2009 119.20 69.10 92.15 13.63 190.80 44.80 118.74 38.44 71.60 24.30
6 6/4/2009 6/8/2009 129.20 86.90 101.16 13.91 179.30 7.10 72.70 37.95 50.10 79.80
7 6/9/2009 6/15/2009 124.90 52.40 97.81 30.22 176.40 50.20 104.93 29.29 51.50 2.20
8 7/14/2009 7/20/2009 133.00 70.40 104.91 17.58 320.70 51.50 153.83 81.25 187.70 18.90
9 7/23/2009 7/27/2009 122.60 49.40 98.34 14.27 307.40 50.40 144.24 68.86 184.80 1.00
10 8/7/2009 8/11/2009 122.00 77.50 93.18 12.79 272.80 38.50 118.59 65.16 150.80 39.00
11 8/19/2009 8/24/2009 126.87 51.81 263.00 64.20 98.34 13.87 130.40 36.90 28.53 37.94
12 9/3/2009 9/9/2009 130.00 97.50 107.75 9.43 192.50 88.90 124.14 34.45 62.50 8.60
13 10/7/2009 10/14/2009 129.70 27.60 85.70 29.37 237.20 27.50 109.05 36.72 107.50 0.10
14 11/6/2009 11/20/2009 194.90 96.40 148.02 19.82 225.60 59.70 108.41 44.32 30.70 36.70

Table 3: Average DO Concentration by Deployment for the ATS System

pH
Upper Lower Summary
Deployment # Deployed Retrieved Max Min S.D. Max Min S.D. xmin xmax
1 3/9/2009 3/12/2009 8.61 6.97 7.60 0.10 9.32 7.71 8.12 0.44 0.71 0.74
2 3/31/2009 4/6/2009 7.77 7.42 7.53 0.06 9.30 7.57 8.16 0.57 1.53 0.15
3 4/14/2009 4/20/2009 7.68 7.33 7.52 0.06 9.30 7.48 8.06 0.41 1.62 0.15
4 4/21/2009 4/28/2009 7.85 7.30 7.56 0.10 10.84 4.98 8.27 0.81 2.99 2.32
5 5/18/2009 5/26/2009 10.00 7.31 7.55 0.23 9.89 7.30 8.13 0.62 0.11 0.01
6 6/4/2009 6/8/2009 7.95 7.48 7.62 0.14 9.01 7.38 7.99 0.51 1.06 0.10
7 6/9/2009 6/15/2009 8.06 7.23 7.54 0.13 10.25 7.37 8.24 0.70 2.19 0.14
8 7/14/2009 7/20/2009 8.26 7.44 7.76 0.24 10.41 7.46 8.56 0.95 2.15 0.02
9 7/23/2009 7/27/2009 8.09 7.47 7.67 0.17 9.74 7.48 8.42 0.83 1.65 0.01
10 8/7/2009 8/11/2009 8.50 7.40 7.71 0.24 9.78 7.43 8.14 0.79 1.28 0.03
11 8/19/2009 8/24/2009 7.96 7.38 7.61 0.16 10.43 7.57 8.31 0.72 2.47 0.19
12 9/3/2009 9/9/2009 7.82 7.32 7.49 0.12 9.06 7.43 8.03 0.48 1.24 0.11
13 10/7/2009 10/14/2009 7.90 7.24 7.42 0.11 10.55 7.10 8.25 0.80 2.65 0.14
14 11/6/2009 11/20/2009 7.90 7.18 7.50 0.12 9.39 7.16 7.94 0.65 1.49 0.02

Table 4: Average pH Data by Deployment for the ATS System

93
 Specific Conductivity
Upper Lower Summary
Deployment
Deployed Retrieved Max Min S.D. Max Min S.D. xmin xmax
#
1 3/9/2009 3/12/2009 657.00 528.00 605.03 20.91 634.00 449.00 603.51 19.71 23.00 79.00
2 3/31/2009 4/6/2009 498.00 388.00 457.07 29.22 738.00 381.00 445.40 30.67 240.00 7.00
3 4/14/2009 4/20/2009 520.00 238.00 413.14 71.98 516.00 205.00 407.67 70.45 4.00 33.00
4 4/21/2009 4/28/2009 542.00 449.00 500.49 22.26 573.00 405.00 470.95 28.23 31.00 44.00
5 5/18/2009 5/26/2009 515.00 95.00 425.98 72.76 521.00 155.00 382.36 122.49 6.00 60.00
6 6/4/2009 6/8/2009 606.00 512.00 574.77 18.90 602.00 487.00 558.57 26.36 4.00 25.00
7 6/9/2009 6/15/2009 631.00 181.00 425.34 131.22 626.00 183.00 430.62 120.72 5.00 2.00
8 7/14/2009 7/20/2009 630.00 508.00 587.02 19.80 597.00 55.00 555.50 59.03 33.00 453.00
9 7/23/2009 7/27/2009 567.00 6.00 543.70 27.61 585.00 292.00 500.54 60.36 18.00 286.00
10 8/7/2009 8/11/2009 552.00 3.00 411.46 173.26 590.00 2.00 403.72 162.39 38.00 1.00
11 8/19/2009 8/24/2009 683.00 432.00 598.35 51.38 678.00 433.00 599.70 49.02 5.00 1.00
12 9/3/2009 9/9/2009 522.00 371.00 447.40 0.12 526.00 360.00 448.47 37.51 4.00 11.00
13 10/7/2009 10/14/2009 602.00 199.00 372.73 106.50 590.00 90.00 351.05 124.28 12.00 109.00
14 11/6/2009 11/20/2009 634.00 266.00 562.22 50.77 630.00 275.00 554.84 48.03 4.00 9.00

Table 5: Average Specific Conductivity Data for the ATS System

The total nitrogen and total phosphorus concentrations at the inflow and outflow, along with the percent
removal, from each sampling event are given in Tables 6 and 7.

Average
Average Average Average
Concentration %
Date Concentration at % Removal Date Concentration Concentration at
at Outflow Removal
Inflow (ppm) at Inflow (ppm) Outflow (ppm)
(ppm)

5/19/2009 3.76 3.71 1.4% 7/20/2009 14.74 18.68 -26.7%

5/28/2009 3.72 4.13 -11.0% 7/22/2009 18.03 12.94 28.2%

6/9/2009 2.96 2.29 22.5% 7/23/2009 22.83 12.19 46.6%

6/16/2009 2.48 2.18 12.2% 7/24/2009 20.55 20.52 0.1%

7/11/2009 3.91 3.78 3.4% 7/25/2009 12.81 5.94 53.6%

7/14/2009 3.69 3.37 8.8% 7/26/2009 13.14 6.97 46.9%

7/15/2009 3.43 2.63 23.3% 7/27/2009 15.13 15.73 -4.0%

7/16/2009 3.74 3.74 0.0% 8/7/2009 22.50 13.30 40.9%

94
7/17/2009 5.36 3.76 29.9% 8/18/2009 6.16 4.13 33.0%

7/18/2009 3.66 3.27 10.5% 8/31/2009 7.51 12.21 -62.7%

7/19/2009 3.10 3.34 -7.5%

Table 6: Total Nitrogen Concentrations Above and Below the ATS System, with Percent Removals

95
 Average Average Average
Average
Concentration Concentration Concentration %
Date Concentration at % Removal Date
at Outflow at Inflow at Outflow Removal
Inflow (ppm)
(ppm) (ppm) (ppm)

5/19/2009 0.211 0.170 19% 7/19/2009 0.082 0.013 84%

5/28/2009 0.252 0.248 2% 7/20/2009 0.108 0.046 58%

6/9/2009 0.172 0.142 17% 8/4/2009 0.319 0.192 40%

6/16/2009 0.088 0.053 40% 8/24/2009 0.225 0.077 66%

7/14/2009 0.200 0.047 77% 9/9/2009 0.295 0.101 66%

7/15/2009 0.148 0.074 50% 9/14/2009 0.313 0.183 42%

7/16/2009 0.024 0.005 79% 10/21/2009 0.104 0.137 -31%

7/17/2009 0.049 0.005 89% 11/9/2009 0.191 0.111 42%

7/18/2009 0.023 0.007 69%

Table 7: Total Phosphorus Concentrations Above and Below the ATS System, with Percent Removals

Figure 5 and Figure 6 show the removal rates of total phosphorus and total nitrogen, respectively.
Between March and mid-May, removal rates were relatively low, and water quality data indicated on one
occasion that phosphorus increased across the system. Phosphorus removal rates began accelerating in
mid June, and measurements taken in July show removal of over 80%. Orthophosphate and nitrate-nitrite
data can be found in Appendix A.

100%
Total Phosphorus Percent Removal

80%

60%

40%

20%

0%

-20%

-40% 96
4/12/2009 5/13/2009 6/13/2009 7/14/2009 8/14/2009 9/14/2009 10/15/2009 11/15/2009

Figure 5: Total Phosphorus Percent Removal from the ATS System

 60.0%

40.0%
Total Nitrogen Removal

20.0%

0.0%

-20.0%

-40.0%

-60.0%

-80.0%
5/7/2009 5/27/2009 6/16/2009 7/6/2009 7/26/2009 8/15/2009 9/4/2009 9/24/2009

Figure 6: Total Nitrogen Percent Removal from the ATS System

In addition to the single water quality sampling events during each harvest period, an extended water
sampling regiment was implemented in July. Water samples were taken every day for 14 days to
determine the variations in nutrient removal throughout multiple harvest periods. Figure 7 shows the
daily change in phosphate and total nitrogen removal from the harvest period from July 11th through July
26th.

100.0%
80.0%
60.0%
40.0%
20.0%
0.0%
-20.0%
-40.0%
7/11/2009
7/12/2009
7/13/2009
7/14/2009
7/15/2009
7/16/2009
7/17/2009
7/18/2009
7/19/2009
7/20/2009
7/21/2009
7/22/2009
7/23/2009
7/24/2009
7/25/2009
7/26/2009

Orthophosphate Removal Total Nitrogen Removal

97
Figure 7: Daily Orthophosphate Removal Across Two Harvest Periods at the ATS System
WEATHER DATA

Figures 8 and 9 show precipitation and ambient temperature data, respectively, over the course of system
operation.

3.5

2.5
Precipitation (in.)

1.5

0.5

Figure 8: 2009 Precipitation Data Collected Daily for Springdale, AR

100
90
80
Temperature (F)

70
60
50
40
30
20
10

Maximum Temperature Minimum Temperature

98
Figure 9: 2009 Maximum and Minimum Ambient Temperature Data Collected Daily for Springdale, AR
 ANALYSIS
Figure 10 compares the rates of removal of orthophosphate to biomass growth during the first five months
of system operation. Figure 11 compares nitrate-nitrite removal to algal growth. It is important to note
that the dip in biomass production in May was the result of a system shutdown early in that month.

60.00 100%
Biomass Production (g/m2/day)

80%

Orthophosphate Removal
50.00 60%
40.00 40%
20%
30.00 0%
-20%
20.00 -40%
10.00 -60%
-80%
0.00 -100%
3/1/20093/31/20094/30/20095/30/20096/29/20097/29/20098/28/20099/27/200910/27/2009

Biomass Production Orthophosphate Removal

Figure 10: Biomass Production and Orthophosphate Removal Over the 9-month Harvesting Period at the ATS

60.00 80%
Biomass Production (g/m2/day)

50.00 60%

Nitrate-Nitrite Removal
40%
40.00
20%
30.00
0%
20.00
-20%
10.00 -40%

0.00 -60%
3/1/2009 4/10/2009 5/20/2009 6/29/2009 8/8/2009 9/17/2009 10/27/2009

Biomass Production Nitrate-Nitrite Removal

Figure 11: Biomass Production and Nitrate-Nitrite Removal Over the 9-month Harvesting Period at the ATS

99
Figure 12 compares the removal of total phosphorus to biomass production from May 2009 through
November 2009. Figure 13 compares total nitrogen removal to biomass production from May 2009
through August 2009.

60.00 100%
Biomass Production (g/m2/day)

80%

Total Phosphorus Removal

50.00
60%
40.00
40%
30.00
20%
20.00
0%
10.00 -20%

0.00 -40%
5/20/2009 6/29/2009 8/8/2009 9/17/2009 10/27/2009

Biomass Production Total Phosphorus Removal

Figure 12: Biomass Production and Total Phosphorus Removal Over the 9-month Harvesting Period at the ATS

60.00 60%
Biomass Production (g/m2/day)

50.00 40%

Total Nitrogen Removal

20%
40.00
0%
30.00
-20%
20.00
-40%
10.00 -60%

0.00 -80%
5/19/2009 6/28/2009 8/7/2009

Biomass Production Total Nitrogen Removal

Figure 13: Biomass Production and Total Nitrogen Removal Over the 9-month Harvesting Period at the ATS

100
Figures 14 and 15 compare biomass production to maximum temperature and average precipitation,
respectively.

60.00 120
Biomass Production (g/m2/day)

50.00 100

Temperature (F)
40.00 80

30.00 60

20.00 40

10.00 20

0.00 0
3/1/2009 4/20/2009 6/9/2009 7/29/2009 9/17/2009 11/6/2009

Biomass Production Maximum Temperature

Figure 14: Biomass Production and Maximum Ambient Temperature Over the 9-month Harvesting Period at the ATS

60.00 3.5

3
Biomass Production (g/m2/day)

50.00

2.5

Precipitation (in.)
40.00
2
30.00
1.5
20.00
1

10.00 0.5

0.00 0
3/1/2009 4/20/2009 6/9/2009 7/29/2009 9/17/2009 11/6/2009

Biomass Production Precipitation

Figure 15: Biomass Production and Average Precipitation Over the 9-month Harvesting Period at the ATS

101
 SUMMARY AND CONCLUSIONS

The average rate of biomass production at the ATS for the first nine months of system operation was
25.30 g dry weight/m2/day. The six period moving averages shown in Figure 1 shows the shift in
production levels from the startup phase. Productivity levels from the first three months of operation
averaged 15 g/m2/day, while the second and third three-month period averaged production rates of 32
g/m2/day and 22 g/m2/per day, respectively.

The ATS system showed high rates of removal of phosphorus, but lower removal rates for nitrogen. The
removal for orthophosphate and total phosphorus were 40% and 48%, respectively. The removal rates for
nitrate-nitrite and total nitrogen, however, were 13% and 12%. Dissolved oxygen levels at the outflow
reached up to 300% saturation, while pH values over 10 were observed.

In general, productivity was heavily impacted by weather and system shutdowns. Freezing in the trough
was observed in late February and early March, as well as in December of 2009 (see Appendix B for
images). During these times, very little growth was observed. In addition to freezing, there were several
instances throughout the year in which Northwestern Arkansas experienced heightened levels of
precipitation. During May, early August, and early and late October, the area was inundated, and a
decrease in productivity was observed. In addition, system shutdowns interrupted production. The most
drastic shutdown occurred in early May and lead to a complete loss of the algal turf established in March
and April.

102
 APPENDIX A: ANION DATA
Table 1A shows the orthophosphate concentrations at the inflow and outflow, as well as the percent
orthophosphate removed. It is important to note that, due to a matrix interference, the magnitude of the
concentrations at the inflow and outflow is inaccurate. However, the concentrations relative to one
another, and thus, the percent removal of the orthophosphate, is considered to be accurate. The percent
removal of orthophosphate over time is given in Figure 1A.

Average Average
Amount
Date Concentration at Concentration at % Removal
Removed (ppb)
Inflow (ppb) Outflow (ppb)

3/3/2009 110.28 206.75 -96.47 -87.48%

3/9/2009 234.84 130.98 103.86 44.22%

3/20/2009 213.73 197.21 16.51 7.73%

4/8/2009 406.56 375.56 31.01 7.63%

4/14/2009 224.63 243.68 -19.05 -8.48%

4/28/2009 989.84 914.02 75.82 7.66%

5/19/2009 1092.46 937.88 154.59 14.15%

5/28/2009 1025.71 980.28 45.44 4.43%

6/9/2009 897.98 710.79 187.18 20.85%

6/16/2009 739.14 484.79 254.35 34.41%

6/28/2009 233.58 28.22%

6/29/2009 191.21 49.52%

6/30/2009 370.13 69.84%

7/2/2009 237.43 89.52%

7/7/2009 611.56 58.18%

7/8/2009 403.42 49.40%

7/11/2009 526.24 345.55 180.69 34.34%

7/14/2009 861.46 451.15 410.31 47.63%

7/15/2009 713.16 322.54 390.62 54.77%

103
7/16/2009 555.99 135.05 420.94 75.71%

7/17/2009 520.23 87.00 433.23 83.28%

7/18/2009 520.23 84.87 435.36 83.69%

7/19/2009 498.48 153.08 345.40 69.29%

7/20/2009 990.97 491.57 499.40 50.40%

7/22/2009 494.99 133.22 361.78 73.09%

7/23/2009 705.62 251.99 453.64 64.29%

7/24/2009 469.27 198.78 270.50 57.64%

7/25/2009 570.16 173.64 396.53 69.55%

7/26/2009 540.33 264.10 276.23 51.12%

7/27/2009 1164.94 1080.46 84.48 7.25%

8/4/2009 881.37 640.54 240.83 27.32%

8/7/2009 249.76 76.74 173.02 69.27%

8/18/2009 673.56 465.10 208.46 30.95%

8/24/2009 848.4003 545.6117 302.79 35.69%

8/31/2009 937.0577 469.3637 467.69 49.91%

9/9/2009 838.8460 583.4263 255.42 30.45%

9/14/2009 793.5697 516.5417 277.03 34.91%

10/21/2009 586.0293 376.7233 209.31 35.72%

11/9/2009 553.6933 440.2010 113.49 20.50%

104
 Figure 1A: ATS Orthophosphate Concentrations from March 2009 to November 2009, with Percent Removal.

100.00%
80.00%
60.00%
Orthophosphate Removal

40.00%
20.00%
0.00%
-20.00%
-40.00%
-60.00%
-80.00%
-100.00%
2/6/2009 3/28/2009 5/17/2009 7/6/2009 8/25/2009 10/14/2009 12/3/2009

Figure 1A: Orthophosphate Concentrations Over the 9-month Harvesting Period at the ATS, with Percent Removal

105
Figure 2A shows the nitrate-nitrite concentrations at the inflow and outflow, as well as the percent nitrate-
nitrite removed. Figure 2A shows the percent removal of nitrate-nitrite over time.

Average Average
Amount
Date Concentration at Concentration at % Removal
Removed (ppb)
Inflow (ppb) Outflow (ppb)

3/3/2009 3870.4140 3231.3890 639.0250 16.51%

3/9/2009 2891.7950 2993.4460 -101.6510 -3.52%

3/20/2009 3791.9740 3283.9450 508.0290 13.40%

4/8/2009 4248.049667 3994.115 253.9346667 5.98%

4/14/2009 3197.645667 2888.047333 309.5983333 9.68%

4/28/2009 2848.008 2731.989 116.019 4.07%

5/19/2009 2157.127 3052.351 -895.224 -41.50%

5/28/2009 2968.645 2871.658333 96.98666667 3.27%

6/9/2009 2240.4415 1730.1265 510.315 22.78%

6/16/2009 1812.981 1484.305333 328.6756667 18.13%

7/11/2009 3066.595 2741.041 325.554 10.62%

7/14/2009 2985.298 2596.637667 388.6603333 13.02%

7/15/2009 2500.595 1918.711667 581.8833333 23.27%

7/16/2009 5920.33 2837.221333 3083.108667 52.08%

7/17/2009 7774.538333 3750.443333 4024.095 51.76%

7/15/2009 5491.467 2017.865 3473.602 63.25%

7/19/2009 2181.951333 1992.118333 189.833 8.70%

7/26/2009 2760.655333 2809.005333 -48.35 -1.75%

7/27/2009 2595.24 2204.283333 390.9566667 15.06%

8/4/2009 4274.074667 3933.128667 340.946 7.98%

8/7/2009 3485.407667 3382.344667 103.063 2.96%

8/18/2009 2918.984333 2787.770333 131.214 4.50%

106
 8/24/2009 3642.742 3093.269667 549.4723333 15.08%

8/31/2009 3714.199333 3065.861667 648.3376667 17.46%

9/9/2009 3737.735 3337.911333 399.8236667 10.70%

9/14/2009 3239.755667 3239.755667 0 0.00%

10/21/2009 3969.486 3827.263667 142.2223333 3.58%

11/9/2009 3542.023333 3409.298333 132.725 3.75%

Figure 2A: Nitrate-nitrite concentrations from March 2009 to November 2009, with Percent Removal

107
 80.00%

Nitrate-Nitrite Removal 60.00%

40.00%

20.00%

0.00%

-20.00%

-40.00%

-60.00%
2/6/2009 3/28/2009 5/17/2009 7/6/2009 8/25/2009 10/14/2009 12/3/2009

Figure 2A: Nitrate-Nitrite Removal Over the 9-month Harvesting Period at the ATS

108
 APPENDIX B: PICTURES

Algal Turf Scrubber System in Springdale, AR Harvesting at the ATS System

Spring Creek in Springdale, AR, at ATS Inflow Algae Samples Harvested from the Lower,
Site Middle, and Upper Sections of the ATS System

109
Dried Algae Samples from ATS Harvest. Green Water Filtering Analysis for Dry
Samples Dried at 105C Weight. Samples Dried at 105C

ATS Growth Medium in February 2009 ATS Growth Medium in March 2009

110
ATS Growth Medium in April 2009 ATS Growth Medium in May 2009

ATS Growth Medium in June 2009 ATS Growth Medium in July 2009

ATS Growth Medium in August 2009 ATS Growth Medium in September 2009

111
 ATS Growth Medium in October 2009 ATS Growth Medium in November 2009

ATS System Freezing Event in February 2009 ATS System Freezing Event in December 2009

Algal Analyses for the Springdale, AR. ATS

The ATS system found in Springdale, AR was sampled and harvested over the period of the study by
Heather Sandefur, and samples of the mesh were sent to NMNH for qualitative and quantitative
analyses. The analyses were carried out as described in the Algae of the ATS Systems portion of this
report. Following are charts demonstrating the relative abundance of the different algal groups
throughout the study.

112
 ATS Springdale, AR 3/17/2009
Chlorophyta
Cyanobacteria 1% Euglenophyta*
1% 0%

Bacillariophyta
98%

Diatoms account for 98% of the algal biomass, followed by chlorophytes and cyanobacteria,
respectively. The most dominant diatoms were chains of Diatoma vulgare. The chlorophytes are
extremely low; however, the system has only been running for a short period and should come
up to speed shortly. Note: diatoms were the first to colonize this system, after the high P/N
levels were removed (fertilizer). Previously, Phormidium spp. and other cyanobacteria
dominated the system. *1 cell of Euglena was encountered in the sample.

ATS_Springdale, AR 4/11/2009
Cyanophyta
5% Chlorophyta
1%

Bacillariophyta
94%

Diatoms account for 94% of the algal biomass, followed by cyanobacteria and chlorophytes,
respectively. The most dominant filamentous diatoms were chains of Diatoma spp. and some
Melosira varians. Gomphonema, Navicula, and Nitzschia were also abundant. The chlorophytes
are low; however, there are more filaments than the previous sample. Surprisingly, many
Phormidium filaments were found on the mesh section.

113
 ATS Springdale, AR 4/16/2009
Cyanobacteria
4% Chlorophyta
10%

Bacillariophyta
86%

Diatoms still account for 86% of the algal biomass, followed by chlorophytes and cyanobacteria
respectively. The most dominant filamentous diatoms were chains of Diatoma spp. and Melosira
varians. Gomphonema, Navicula, and Nitzschia were also abundant. Filamentous greens, like
Stigeoclonium, cf. Uronema, Mougeotia, as well as coccoidal greens are on the floway.
Phormidium filaments were not as abundant as the previous harvest.

ATS Springdale. July 2009

Cyanobacteria
3%

Chlorophyta
37%
Bacillariophyta
60%

Diatoms account for 60 % of the algal biomass, followed by chlorophytes (37%) and
cyanobacteria (3%), respectively. This is extremely different than the April harvest which showed
a dominance of 94% diatoms. The most dominant filamentous diatoms are chains of Melosira
varians and Fragilaria sp. This also is different than April diatoms which were dominated by
Diatoma spp. Gomphonema, Navicula, and Nitzschia continue abundant. Chlorophytes are
mainly dominated by a filamentous Ulothrix sp. and a pseudoparenchymatous Coleochaete sp.,
there are more filaments than the previous sample.

114
 ATS Sprindale. Aug 8 2009
Cyanobacteria
1%

Chlorophyta
35%
Bacillariophyta
64%

Diatoms account for 64 % of the algal biomass, followed by chlorophytes (35%) and
cyanobacteria (1%), respectively. The most dominant diatoms are species of Synedra, Diadesmis,
Fragilaria, and Navicula; Melosira has decreased. The chlorophytes consist mainly of
chlorrococcalean species like Scenedesmus and Desmodesmus, which are able to get caught in
the filaments of Cladophora and Oedogonium, present on the floway.

115
 Appendix D

Biofuel Production from Algal Turf

A Report

On

A Project to Process Algal Turfs

From ATS systems to

Biofuels and other by-products

John Miller, Stephen Bertman

Chemistry Department

Sarah Hill

Anthropology Department

Western Michigan University

Kalamazoo, Michigan, December, 2009

116
Biofuel Production from Algal Turf
Reporting period through 31 October 2009

This report is an overall summary of the result of the project, supported by the Smithsonian Institution
and conducted at Western Michigan University (WMU), to examine the prospects of producing biofuels
from algal biomass, particularly the mixed benthic communities grown as an algal turf.

The contents of this report are focused on:

1. Processing requirements for efficient extraction of biofuel feedstocks.

2. Classification and quantitation of the algal biofuel feedstocks (oils and carbohydrates)
3. Conversion of the algal biofuel feedstocks into biofuels
4. Potential for non-fuel components in algal biomass
5. Biofuel yields and energy potential estimates

PROCESSING REQUIREMENTS FOR EFFICIENT EXTRACTION OF BIOFUEL

FEEDSTOCKS.

Biomass processing
Processing the raw algal biomass into biofuel feedstocks, and ultimately into biofuels (ethanol, butanol,
biodiesel) or other value-added products was a primary goal of the project. Extensive experimentation
with a range of biomass pre-processing techniques led to the process flow diagram shown in Figure 1.
The dry-mass balances are shown after each separation separation step. Some operations have not
been fully down-selected, and all need to be optimized for scales larger than about a kilogram (dry
mass).

Figure 1. Process flow

diagram for algal biomass processing into biofuels. (Percentages show dry-mass balances.)

117
 Sample handling
Algal turf biomass was primarily obtained from collaborators at a range of sites shown in Table 1.
Initially, algal samples were shipped to WMU either frozen or dried, as decomposition was thought to be
a concern when shipping fresh biomass. However, decomposition was found to be minimal so long as
the biomass was well-drained, kept as cool as possible, wrapped to exclude air, and shipped quickly.
Freezing the whole biomass tended to cause separation of the cellular solids and the intrinsic water.
Moreover, shipping cryogens or cold-packs increases the complexity and expense of transport. Thus
shipping frozen was effectively abandoned.

Table 1. Algal turf sources

Collaborator Site Comments

Hydromentia Taylor Creek, FL Frozen samples

Hydromentia Powell Creek, FL Fresh samples

Single sample. Very high

Hydromentia STA-1W, FL
carbonate

Dr. Pat Kangas (U. Md) Muddy Run, PA Frozen and fresh samples

VA Institute of Marine Sciences,

Paul Richardson Fresh samples
VA

Dr. Walter Mulbry USDA, MD Single dried sample

Drying was generally considered inappropriate for shipping biomass, since thermal drying could change
the chemistry of the biomass components and potentially cause loss of more-volatile components, while
freeze-drying large quantities of very wet material is generally impractical. Furthermore, the ultimate
goal is to produce biofuel from fresh, whole biomass, so dried materials were considered somewhat
unrepresentative from a processing standpoint, even if the overall chemistry was likely preserved.

Sample Storage
Preservation of the algal samples for archival purposed posed a bit of a problem. The wet biomass
tends to support growth of a wide range of common molds and bacteria. Simple freezing of the
homogenized biomass is effective over the short term, but the materials tend to dehydrate (freezer-
burn) quickly. Thus, aliquots of most of the homogenized biomass samples were lyophilized for long-
term storage.

118
CLASSIFICATION AND QUANTITATION OF THE ALGAL BIOFUEL FEEDSTOCKS (OILS
AND CARBOHYDRATES)

Oils
The total quantity of oil available in the algal turf samples was determined by solvent extraction after
cell disruption. Oil yield data are shown in Figure 2. The variability between sources is generally low
relative to the sample-to-sample variation from the same source. Overall, the oil yield is generally low,
about 3-6 wt%. However, there are some high-value components in the algal oil.

9
8
7
Oil extracted / wt%

6
5
4
3
2
1
0

Figure 2. The oil content of multiple algal turf samples from several sources as determined by
solvent extraction. The error bars represent the standard deviation for three or more replicate
extractions.
The fatty acids in the oils were characterized by converting all of the acids to their corresponding methyl
esters the so-called FAMEs analysis and analyzing by gas chromatography, for routine quantitation,
and by gas chromatography/mass spectrometry, for compound identification. The results are
summarized in Figure 3, where the oils are grouped by chain length, C14-C20. The longer chain length
oils are particularly interesting, as they largely comprise high-value polyunsaturated fatty acids (PUFAs);
about 30% of the C20:x (where x indicates degree of unsaturation) is eicosapentaenoic acid (EPA, or
C20:5), a common component in fish-oil dietary supplements. Other high-value PUFAs are also present,
including docosahexaenoic acid (DHA, or C22:x).

119
 100%
90%
80% PUFAs
70%
60%
50% C20:x
40%
C18:x
30%
20% C16:x
10% C14:x
0%

Figure 3. The relative fatty acid compositions of extracted algal oils as determined from
FAMEs analysis.

Carbohydrates
The carbohydrate content of the algal turf biomass from different sources was analyzed for
carbohydrate content and composition by the sequence below.

1. Soluble carbohydrates -- Exhaustive Soxhlet extraction with 80% ethanol to determine soluble
carbohydrate by gravimetry.
2. Insoluble -amyloses -- Enzymatic treatment of the residual material with -amylase, followed
by Soxhlet extraction and gravimetry.
3. Celluloses -- Enzymatic treatment of the residual material with cellulase, followed by Soxhlet
extraction and gravimetry.
A comparison of carbohydrate compositions for different algal sources is shown in Figure 4. This
analytical sequence is quite time consuming, so not all biomass samples were subjected to the entire
protocol. However, the quantites of soluble carbohydrates were similar for most of the remaining
samples that were tested. Thus we infer that the total carbohydrate content was in the same range as
these sources. In general it appears that the carbohydrate content of algal turf is sufficiently high to
merit consideration as a feedstock for bio-alcohol fuels.

120
 60%

50%

40% Amylase-labile
Weight %

carbohydrate
30% Cellulase-labile
Carbohydrate
20%
Soluble Carbohydrate

10%

0%
Taylor Creek Muddy Run

Figure 4. Comparison of carbohydrate compositions from different algal turf sources.

In one instance, we were able to analyze the carbohydrate composition as a function of distance down
the first Muddy Run floway. Those data are shown in Figure 5. Overall, the proportion of soluble
carbohydrates and those that are digested by cellulase (celluloses and hemi-celluloses) do not show a
stong dependence with floway position, while those that are digested by -amylase (starches and other
amyloses) generally appear to increase with distance down the floway. Additional studies of this type
will be necessary to determine if this is a general trend.

70%

60%

50%
Weight %

40% Amylase-labile
carbohydrate
30% Cellulase-labile
20% Carbohydrate
Soluble Carbohydrate
10%

0%
6 21 30 43
Distance down Floway / m

Figure 5. Dependence of carbohydrate composition on position in an ATS floway for algae

grown at Muddy Run, PA. (The error bars represent the standard deviation of at least three
replicate measurements.)

121
CONVERSION OF THE ALGAL BIOFUEL FEEDSTOCKS INTO BIOFUELS
Carbohydrates into butanol
No effort was made at WMU to convert algal carbohydrates into butanol. However several batches of
carbohydrate extract were shipped to Dr. Jamie Hestekin of the University of Arkansas for butanol
fermentation using a modified Ramey process.

Carbohydrates into Ethanol

The results of fermentation experiments using yeast (Sacchromyces cerevisiae) to produce ethanol from
algal turf carbohydrate extracts are shown in Figure 6. Although ethanol was generated, the yield was
far below what was expected based on the total carbohydrate content.

Our initial hypothesis was that the unmodified algal carbohydrate extracts were poor substrates for the
yeast, which is most effective at fermenting simple sugars. Thus we carried out a series of experiments
digesting the homogenized and acid-processed algal turf with several carbohydrate-digesting enzymes,
selected based on the reported carbohydrate species in algae, including starches, pectins, agars, (hemi-
)celluloses, and alginates. Analysis of the free glucose in the algal biomass before and after enzymatic
treatment is shown in Figure 7. Clearly the increase in free glucose is not dramatic. Nevertheless, the
slight increases in ethanol yield (Figure 6) using enzyme treated algae are consistent with the small
measured increases in free glucose.

Although these ethanol and free-glucose yields were somewhat discouraging initially, we suspected that
the algal carbohydrates could be very complex, and that no single enzyme would be adequate to
convert them to an appropriate substrate for the yeast. Thus we prepared a cocktail consisting of
several enzymes. Preliminary results (also in Figure 6) suggest that algal turf carbohydrates digested
with such a mixed-enzyme regime can be used for yeast fermentation into ethanol. Additional work will
be necessary to identify the optimum enzyme mixture, and to determine if the balance of its
components might need to be adjusted based on the origin or composition of the algal biomass.

122
 3.0

2.5

2.0
[EtOH] (vol%)

1.5

1.0

0.5

0.0
amylase

alginate

"cocktail"
pectinase

5% sucrose
SWE only
ph3 and

enzyme
lyase

Figure 6. The yield of ethanol from a six-day yeast fermentation of 5 wt% algal carbohydrate
solutions extracted using subcritical water extraction (pH=3, 125C, ca. 2 bar) and subsequent
treatment with various enzymes. For reference, the six-day ethanol yield from a 5 wt% sucrose
solution is also shown.

0.025

0.020
[Glucose] (mg/mL)

0.015

0.010

0.005

0.000
Control Cellulase a-Amylase Pectinase

Figure 7. Assays of free glucose after treating homogenized algal turf with individual
carbohydrate-digesting enzymes. The error bars represent the standard deviation for at least
three replicate measurements.
123
An alternative hypothesis was that some component in the biomass might be an inhibitor of ethanol
formation. Solid-phase micro extraction (SPME) headspace analysis of the algal turf extracts did show
some known yeast inhibitors, notably benzoic acid. However, preliminary fermentation experiments
suggest that addition of algal turf carbohydrate does not inhibit, and may even promote, sucrose
fermentations. More work will be necessary to fully assess this hypothesis.

Oils into biodiesel

The oils that were extracted from the algal turf biomass are readily converted to biodiesel, the methyl
esters of the fatty acids in the oils. The biodiesel syntheses were carried out at the laboratory scale, by
now-standard transesterification chemistry, whether base-catalyzed (NaOH or KOH) or Lewis-acid
catalyzed (BF3 or AlCl3) conditions, and gave overall reaction yields that were comparable to those using
terrestrial vegetable oils. (The FAMEs analyses reported above are examples of quantitative conversion
chemistry.) Given the inherently low proportion of oil and the relatively small size and small number of
algal biomass samples extracted, we were unable to investigate biodiesel synthesis on anything larger
than about gram-scale during this project.

POTENTIAL FOR NON-FUEL COMPONENTS IN ALGAL BIOMASS

A large proportion (~50-60%) of the algal turf biomass is not appropriate for conversion into biofuels.
Nevertheless is still has value. Our current thinking is that the non-fuel feedstock residue is most useful
as a fertilizer. Although there is no free nitrate (NO3), the residual solids contain about 5.5 wt%
nitrogen. In addition preliminary measurements indicate high levels of both phosphorus and potassium.
To assess the potential as fertilizer, we carried out growth experiments using Brassica rapa, a fast-
growing mustard relative. The growth curves from those experiments are shown in Figure 8. Overall,
algal turf residual solids enhanced growth over plain potting soil, and may be comparable to commercial
organic fertilizer. A more in-depth study of this fertilizer potential is merited.

90
80 Figure 8. The growth rates
70 of Brassica Rapa fertilized
"organic" with algal turf solids
60
remaining after removal of
Plant height / mm

50 biofuels feedstocks, as well

Algae Solids
40 A as a commercial organic
fertilizer and no fertilizer
30 (soil).
20
10
0
124
0 10 20 30
Growth time / days
BIOFUEL YIELDS AND ENERGY POTENTIAL ESTIMATES
We can combine the biomass analyses and the biofuel output yields to create an estimate of the overall
biofuel energy potentials for algal turf. Those estimates are included in Table 1. In summary, algal turf
is likely a viable source for biofuel feedstocks, capable of providing several product streams.

Table 2. Biofuel yields and energy potential

Algal component dry wt% Biofuel or Energy potential Energy potential /
product / kJ g-1 dry wt BTU ton-1 dry wt

Carbohydrates 40% Ethanol or 10.1 9.7 M

butanol

Oils 5% Biodiesel 1.9 1.8 M

PUFAs 0.5% Supplements n/A

Residual Solids ~50% Fertilizer n/A

125
Appendix E

126
The Problem
The waters of the Chesapeake Bay are polluted primarily by nitrogen and phosphorus originating from
agricultural land, urban wastewater plants, and airborne contaminants from vehicles and power plants.
This results in depleted oxygen levels in the Bay, which in turn threaten virtually all macro-organisms
from fish to crabs. Populations of oysters are at an all-time low, and habitats such as underwater grass
beds and wetlands continue to be degraded. This degradation threatens the entire ecology of the Bay,
most noticeably by massive fish kills caused by summer dead zones. This in turn jeopardizes the well
being and livelihood of the 16.6 million residents within the Chesapeake Bay Watershed. The watershed
stretches across 64,000 square miles and encompasses parts of six states DE, MD, NY, PA, VA, WV
and the entire District of Columbia. Threading through the watershed are more than 100,000 streams
and rivers that eventually flow into the Bay. Everyone in the Bay watershed lives within a few minutes of
one of these streams and rivers, which are like pipelines from our communities to the Bay. Most of the
damaging nutrients arrive in the Bay as non-point source, beyond control of previous technology, with
the Susquehanna River itself contributing about half of the detrimental phosphorus levels entering the
Bay.
Executive Order 13508 initiated by President Obama has been issued for government to take a more
active role to protect and restore the health, heritage, natural resources, and social and economic
value of the nations largest estuarine ecosystem and the natural sustainability of its watershed. One
significant element of this order is a 30% reduction of nitrogen and an 8% reduction of phosphorus, over
2008 levels entering the Bay, by the year 2025. The Executive Order recognizes that the efforts of the
past 25 years have not achieved sufficient progress in restoring Bay health.

The Pilot Project

This pilot program will be a large-scale demonstration of the economics of water clean-up combined
with production of biofuel, health and agricultural biproducts. The pilot will include acre-scale
deployment of ATS/AATS systems. ATS systems are already commercially operating at acre scale, and
so part of the project is ready to go to pilot scale. A two-year grant from the Lewis Foundation to the
Smithsonian Institution has allowed a multi-university team to make a bench scale documentation of the
ATS algae to by-products process. An Algae Consortium has been formed consisting of Constellation
Energy, Exelon Corporation, Statoil, Rolls-Royce, Blackrock Energy, Ecological Systems Technology,
HydroMentia, Smithsonian Institution, University of Virginia, Pennsylvania State University, College of
William & Mary, Virginia Institute of Marine Science, University of Arkansas, Western Michigan
University, University of Maryland, Noblis, Oak Ridge National Labs and the Chesapeake Research
Consortium. The pilot program would be initiated at a strategic location- near the mouth of the
Susquehanna River at Constellation Energys C.P. Crane Power Plant located northeast of Baltimore,
Maryland. Additional sites for ATS/AATS facilities would be evaluated and chosen around the Bay to
optimize nitrogen and phosphorous removal. The pilot program encompasses:

At The C.P. Crane Station of Constellation Energy:

-3 acre land-based ATSTM and water-based AATS for growing, harvesting, and processing wild algae
to reduce Bay nutrients and CO2 while injecting oxygen into Bay waters
-site Mobile Refinery Unit (MRU) for biobutanol production
-site demonstration biodiesel and/or jet fuels from production

127
 2 sequestration, contaminant capture and disposal
-firing of processed algae material

At Alternative Sites/locations:
uction of other high value added co-products such as Omega-3 and fertilizer

Evaluate the potential for the algae residue to produce carbon absorbing terra preta soils.

cycle analysis to assess energy efficiencies.

Constellation Energys C. P. Crane Power Plant site provides the synergistic relationship between the
availability of CO2 from flue gas and algal production; and has infrastructure to support both on-land and
in-water aquatic demonstration. The results of the pilot program can be used directly to confirm
nutrient reductions and oxygen increases in Susquehanna River water flowing into the Bay. Also, the
results will be used to confirm the economic benefits of ATS/AATS and site a 100-acre commercial
demonstration project, followed by a deployment strategy to support the entire Executive Order
directives for nitrogen and phosphorous reduction and oxygen improvement in the Bay waters over the
following 6-10 years.

Total Pilot Program Budget (3 Year Program 2010-2012)

Statoil (monetary): $ 3 million
Constellation & Exelon (in-kind): $ 1 million
Subtotal, Private Sector: $ 4 million
Government Funding Request: $16 million
Total Pilot Program: $20 million

Downsides/Risks

-product mix and benefits

nants and unforeseen factors affecting algae growth or feedstock quality
-products

The Benefits (the 3Es)

Environment
nes.
2 sequestration and contaminants capture via algae uptake.

-3 production from algae will remove need to over-harvest Bay fish.

e over billions of years, provides no risk to the ecosystem; as
opposed to escape of designer single-strain algae. Less risk of disease, and greater ability to adjust to
changing environmental conditions.
Restoration and protection of Chesapeake Bay health.
Energy
-to-ethanol methods.
-firing for electric power generation.
128
Economic Development
-off jobs

An Algae Consortium Approach for:

Environmental Remediation of the Chesapeake Bay
With Sustainable Algal Energy Production
Turning Pollution into Fuel and Food
Supplemental Information

Additional Technology Description

Algal Turf Scrubber (ATS)
Developed initially at the Smithsonian, the land-based ATS technology has a 30 year history of
integration with and control of chemical and biological functions in aquatic ecological systems.
Scaled up over the last 15 years to several acre dimensions for tertiary sewage treatment,
aquaculture (fish farms) and remediation of eutrophic creeks and canals the engineered
ecosystem ATS has a successful track record from coast to coast in the US. Thanks primarily to the
Lewis Foundation and Exelon Corporation, ATS test units have now been modified successfully for
use over the entire range of the Chesapeake Bay. The simple and elegant ATS removes nutrients
from degraded coastal and aquatic environments on a large scale. It also supplies large quantities of
pure oxygen to treated waters to directly remove dead zones. Simultaneously it produces an algal
biomass that can be refined to biofuels and advanced food products, while producing a soil
amendment to restore degraded farmland. The algae require no fertilizer to grow, as do traditional
algal bioreactor designs. ATS exploits the nutrients already present in the water column as a
`fertilizer'.
Aquatic Algae Turf Scrubbing/ In-Water System (AATS)
In July 2009, Statoil provided seed capital funding for our Algae Consortium to research the key
technical drivers and/or concepts behind this successful ATS technology and potentially apply them
to researching, designing and ultimately constructing a scaled up method of growing, harvesting
and processing algae in open water. AATS has the potential to drive the cleanup costs down
significantly, and provide much greater flexibility to the overall deployment strategy.
Both of these systems use natural (wild) algae as opposed to single strain designer algae. The
considerable advantages of this approach include: 1) wild algae grows naturally in streams, rivers
and bays and do not have to have costly supporting, culture operations; 2) these algal strains are
already optimized by Nature to grow year-round in whatever location is chosen for remediation; 3)
an algae mix growing on an ATS or AATS is similar to a portfolio of stocks - minimizes risk of disease
and enhances ability to adjust to changing environmental conditions; and 4) provides no risk to the
ecosystem, as opposed to designer algae escape. In short, these systems use local algal weeds
programmed to grow quickly when provided appropriate substrate. Unlike in agricultural
monoculture, these algal weeds provide everything we need for water remediation and bi-
products.

129