Pilot Test and Field Construction of a Funnel-and-Gate Biobarrier at an abandoned Tar Factory Site

Abstract

A full scale funnel-and-gate biobarrier is developed for the removal of tar oil pollutants at an abandoned tar factory site near the city of Offenbach, Germany. Laboratory and on-site column studies were done to determine the operation parameters for microbiological clean-up of the groundwater polluted with 12,000 µg/L BTEX, 4,800 µg/L PAH, and 4,700 µg/L heterocylic compounds. In the laboratory study, a residence time of approx. 70 h proved to be sufficient for aerobic pollutant biodegradation. Up to 180 mg/L hydrogen peroxide (H2O2) were added and did not lead to any toxic effects to the degrading bacteria. The feasibility of the concept was confirmed in the on-site pilot study performed with a sedimentation tank (removal of ferric iron) and two bioreactors. In the bioreactors, >99.5 % of the pollutants were degraded. Biodegradation activity corresponded to a significant increase in numbers of pollutant degrading bacteria. In the bioreactors, a fast dissociation of H2O2 was observed resulting in losses of oxygen and temporary gas clogging. Therefore, a repeated addition of moderate concentrations of H2O2 proved to be more favourable than the addition of high concentrations at a single dosing port. The full scale biobarrier is consisting of three separated bioreactors thus enabling extended control and access to the reactors. The operation of the funnel-and-gate biobarrier started in march 2007, and represents the first biological EC-PRB in Germany.

Full text

Pilot Test and Field Construction of a Funnel-and-Gate Biobarrier
at an abandoned Tar Factory Site
Andreas Tiehm (tiehm@tzw.de), Axel Müller
(Water Technology Center, Karlsruhe, Germany)
Stefan Alt (CDM Consult GmbH, Alsbach, Germany)
Hermann Schad, Rainer Klein (I.M.E.S. GmbH, Amtzell, Germany)
Christian Weingran, Birgit Schmitt-Biegel (HIM-ASG, Biebesheim, Germany)
ABSTRACT: A full scale funnel-and-gate biobarrier is developed for the removal of
tar oil pollutants at an abandoned tar factory site near the city of Offenbach, Germany.
Laboratory and on-site column studies were done to determine the operation parameters for
microbiological clean-up of the groundwater polluted with 12,000 µg/L BTEX, 4,800 µg/L
PAH, and 4,700 µg/L heterocylic compounds. In the laboratory study, a residence time of
approx. 70 h proved to be sufficient for aerobic pollutant biodegradation. Up to 180 mg/L
hydrogen peroxide (H
2
O
2
) were added and did not lead to any toxic effects to the degrading
bacteria. The feasibility of the concept was confirmed in the on-site pilot study performed
with a sedimentation tank (removal of ferric iron) and two bioreactors. In the bioreactors,
>99.5 % of the pollutants were degraded. Biodegradation activity corresponded to a
significant increase in numbers of pollutant degrading bacteria. In the bioreactors, a fast
dissociation of H
2
O
2
was observed resulting in losses of oxygen and temporary gas
clogging. Therefore, a repeated addition of moderate concentrations of H
2
O
2
proved to be
more favourable than the addition of high concentrations at a single dosing port. The full
scale biobarrier is consisting of three separated bioreactors thus enabling extended control
and access to the reactors. The operation of the funnel-and-gate biobarrier started in march
2007, and represents the first biological EC-PRB in Germany.
INTRODUCTION
At an abandoned tar factory site near the city of Offenbach, Germany, the groundwater
is contaminated with tar oil pollutants. A plume of dissolved contaminants extends in the
aquifer from the hot spot area downgradient towards the Main river. As part of the RUBIN
research and development network focussing on permeable reactive barriers (PRB), a
funnel-and-gate system is constructed at the site (fig.1). The concept for water clean-up is
based on a three step process including (i) sedimentation of ferric iron, (ii) aerobic
biodegradation of the aromatic hydrocarbons and heterocyclic compounds, and (iii) a
subsequent zone with granular activated carbon (GAC) removing remaining pollutants.
This paper focusses on the studies leading to the operation and design parameters of the
microbiological in-situ reactor.
Under the strictly anaerobic, i.e. methanogenic, conditions in the groundwater plume at
the site, biodegradation of the pollutants is limited resulting in a long plume. However,
most low molecular weight PAH, the BTEX and heterocyclic compounds – toxic pollutants

often occuring at tar oil polluted sites – are degradable in the presence of oxygen. Also in
the presence of nitrate, biodegradation of some specific tar oil pollutants has been
demonstrated (Tiehm et al. 1997; Tiehm & Schulze, 2004; Sagner et al., 2005).
In the bioreactor, the availability of suitable electron acceptors and nutrients has to be
considered as important factor affecting biodegradation efficiency. Due to the high
pollutant concentration in the groundwater at this site (approx. 20 mg/L tar oil pollutants),
hydrogen peroxide was selected as oxygen carrier due to its higher water solubility as
compared to oxygen. Addionally, nitrate was added as alternative electron acceptor.
Laboratory and pilot scale experiments were done to determine (i) the site-specific
efficiency and sequence of pollutant biodegradation, (ii) the locations and concentrations of
hydrogen peroxide addition, and (iii) the suitable operation parameters of the in-situ
biobarrier.
source zone
Kaiserleistraße
Kaiserleibrücke
Nordring
Goethering
Bornheimer Weg
Main
Pilotgate
Pilot-Funnel
2nd
Gate
Funnel
extension
N
0 100 m
Figure 1: Site map with location of the funnel-and-gate biobarrier
MATERIAL AND METHODS
The laboratory column experiments were performed at groundwater temperature (12-
14
°
C) with groundwater samples taken from well GWM19 up-gradient the projected in-situ
reactor. The groundwater samples were filled into 30 L Teflon bags enclosed in nitrogen
flushed vessels and stored under anaerobic conditions. The laboratory column (∅ 25 cm,

length 100 cm, volume 12,7 L) was filled with silica sand (grain size 1 – 2,5 mm), and the
flow direction was top - down. In order to remove precipitated ferric iron, reversal flushing
was performed discontinuously. H
2
O
2
and nitrate was added at the column inlet. During
column operation a second addition port of H
2
O
2
in the upper part of the column was
installed. Column experiments were carried out for a period of 300 d.
The pilot plant columns were operated on-site in an air-conditioned container. Two
bioreactors (stainless steal columns, filled with silica sand) were operated with groundwater
taken from well GWM19. A sedimentation tank with a lamella separator to remove ferric
iron after the first H
2
O
2
- addition was incorporated in front of the bioreactors. H
2
O
2
(35
Vol.-% H
2
O
2
) was added to the sedimentation tank and the two bioreactors. Nitrate was
added to the first bioreactor. Phosphate (KH
2
PO
4
/ NaHPO
4
- buffer) was dosed
discontinuously at both bioreactors. A GAC column was installed at the bioreactor effluent
to remove remaining pollutants. The experimental set-up is shown in figure 2. The on-site
column experiments were operated for a period of 270 d.
Microbial numbers were determined by the most probable number (MPN) microplate
technique with groundwater samples (Stieber et al., 1994).
Figure 2: Scheme of the experimental set-up of the on-site biosorption system

RESULTS AND DISCUSSION
Laboratory column studies
The microbiological investigations at the Offenbach site revealed that sufficient
numbers of microorganisms capable to degrade BTEX and PAH under aerobic conditions
were present in the contaminated soil and groundwater. Denitrifying bacteria were also
detected. Therefore, additional inoculation of allochthonous microorganism was not
required to establish aerobic and denitrifying consortia in the bioreactor.
The results of the laboratory column studies with groundwater from the site showed
that ethylbenzene and naphthalene were already degraded under O
2
-restricted/denitrifying
conditions. In particular the degradation of ethylbenzene was stimulated already by the
addition of nitrate (20 mg/L). The subsequent addition of increasing amounts of H
2
O
2
up
to a concentration of 180 mg/L resulted in increasing biodegradation of the pollutants.
Methane degradation started after removal of the hydrocarbons (fig. 3). The main pollutants
were eliminated in following sequence: ethylbenzene, naphthalene, m-/p-xylene, benzene,
o-xylene. Also heterocyclic aromatic hydrocarbons such as benzothiophene and benzofuran
were degraded. The increasing consumption of electron acceptors was in correlation with
increasing COD removal.
In the column study, a flow rate between 2 to 4 L/d proved to be suitable for complete
pollutant removal. The corresponding residence time of approx. 36 to 72 h was also
selected for the subsequent field study.
0
0,2
0,4
0,6
0,8
1
1,2
0 100 200 300 400 500
time of operation [d]
c/c
0
[-]
benzene ethylbenzene
m-xylene p-xylene
o-xylene naphthalene
methane
180
150
30
15
40
60
100
0
increasing of H
2
0
2
addition [mg/L]
Figure 3: Effect of increasing H
2
O
2
addition on pollutant biodegradation

Pilot plant field study
The laboratory results were confirmed by a subsequent pilot study performed at the
Offenbach site. The flow rate was adjusted to 22 - 44 L/d resulting in residence times
between 36 to 72 h. In the pilot plant, a ferric iron-sedimentation tank with a lamella
separator was incorporated prior to the biodegradation step in order to avoid clogging
effects by precipitation of ferric iron (up to 25 mg/L ferrous iron in groundwater) and to
avoid degassing of supersaturated gases (methane) in the bioreactor.
The addition of H
2
O
2
into the sedimentation tank resulted also in aerobic
biodegradation of part of the pollutants in this area. However, degradation was most
pronounced in the bioreactors. After an adaptation and operation time of 270 d elimination
of the main contaminants was > 99 % at a residence time of 72 hours (tab.1).
Table 1: Bioreactor elimination performance after an operation period of 270 d
parameter
influent
concentration
bioreactor 1
[µg/L]
effluent
concentration
bioreactor 2
[µg/L]
bioreactor elimination
[% of input concentration]
reactor 1 reactor 2
reactor
1 + 2
naphthalene
2,400
1.2
97
3
99.95
ethylbenzene
550
0.54
94
6
99.9
benzene
5,300
10
89
11
99.8
acenaphthylene
12
0.16
81
18
98.7
toluene
66
0.88
73
26
98.7
acenaphthene
110
1.8
72
27
98.4
m- + p- xylene
1,500
10
68
31
99.3
phenathrene
13
0.078
59
40
99.4
fluorene
26
0.27
58
41
99.0
o- xylene
660
7.0
50
49
98.9
fluoranthene
0.66
< 0.02
11
> 86
97.0
anthracene
1.0
0.092
5
> 86
> 90.8
pyrene
0.40
< 0.02
0
> 95
> 95.0
methane
5,140
2,320
3
52
55
BTEX
8,076
28
82
17
99.6
other AHC
5,588
16
83
16
99.6
PAH w/o NAP
163
2,6
68
30
98.5
COD [mg O
2
/L]
142
37
42
21
63.4
bold = main contaminants

The stimulated aerobic biodegradation was also demonstrated by increasing microbial
numbers in the biobarrier system. The highest density of BTEX- and PAH-degrading
bacteria was observed in the first bioreactor (fig. 2) where most of the pollutants were
eliminated (table 1).
In the on-site bioreators and also during the laboratory bioreactor study, the addition of
high concentrations of H
2
O
2
resulted in oxygen gas bubble formation and a reduction of
hydraulic conductivity, due to the fast dissociation of H
2
O
2
and limited solubility of O
2
.
Therefore, the column experiments revealed that moderate additions of H
2
O
2
at several
dosing ports are favourable as compared to the addition of a high amount of H
2
O
2
at a
single dosing port. Taking into consideration these findings and – in general - the better
ability to control the system, a modular full scale system was developed.
1E+01
1E+02
1E+03
1E+04
1E+05
1E+06
1E+07
after 1. H2O2-
addition
after 2. H2O2-
addition
effluent
cells/mL groundwater
total heterotrophs (aerobic)
BTEX- degraders (aerobic)
2-/ 3-ring PAH- degraders (aerobic)
bioreactor
activated carbon
untreated
water
lamella
separator
Figure 4: Numbers (MPN) of pollutant degrading microorganisms in the original
groundwater and after passage of the lamella separator and the biobarrier components
CONCLUSIONS AND FULL SCALE BIOBARRIER DESIGN
The laboratory experiments and the pilot plant field study confirmed the feasibility of
the biobarrier system. Within the bioreactor 99 % of the influent contaminants were
eliminated at a residence time of 72 h. Due to a fast dissociation of H
2
O
2
, a repeated

addition of moderate concentrations is recommended in order to avoid losses of oxygen and
gas clogging in the biobarrier.
Figure 5: Longitudinal section of the full-scale biobarrier system
In the full scale funnel-and-gate biobarrier (fig. 5) the contaminated water will be
collected from the aquifer through a gravel filter. An open water area is separated from the
gravel through a self-contained steel screen. Within the open water area, the groundwater
flow is focussed to a connecting pipe, where hydrogen peroxide will be added. The
following lamella separator is needed for the sedimentation of precipitated iron. At the
outlet tube from the lamella separator (inlet of the first bioreactor), hydrogen peroxide and
nutrients will be added. Distribution of the water to the entire cross-section of the
bioreactor will be accomplished through an open water area. All bioreactors will be filled
with coarse sand. Between the various bioreactors again hydrogen peroxide and nutrients
will be added at the connection of the open water areas for complete degradation. If
pollutant elimination in the full scale system is not sufficient to reach the compliance
values, activated carbon will be filled into the last chamber to remove the remaining
pollutants.
The modular design of the biobarrier with three separated bioreactors follows the
concept of permeable reactive barriers with extended control (EC-PRB) which enable
access to the reactors in case of malfunction. A successful operation of EC-PRBs has been
reported for Zero Valent Iron (ZVI) and Activated Carbon (AC) barriers (Birke et al.,
2005). The operation of the funnel-and-gate biobarrier started in march 2007, and
represents the first biological EC-PRB in Germany.

ACKNOWLEDGEMENT
The study is part of the RUBIN funding priority. The authors gratefully acknowledge
financial support by the HIM and the German Ministry of Education and Research (BMBF,
grant no 02WR0293).
REFERENCES
BIRKE V., BURMEIER H., DAHMKE A., EBERT M. 2005.
The German permeable reactive barrier (PRB) network „RUBIN“: remporary overall
results and lessons learned after five years of work. In: Uhlmann O., Annokkée G., Arendt
F. (Hrsg.) CONSOIL 2005, Proceedings (CD) of the 9th international FZK/TNO
conference on soil-water systems, Bordeaux, 3-7 Oct. 2005: 2865-2887
“RUBIN” 2007. <http://www.rubin-online.de>
SAGNER A., SCHULZE S, STIEBER M, TIEHM A. 2004.
Occurrence and biodegradation of NSO-heteroaromatic compounds in tar-oil polluted
groundwater. In: Proceedings (CD) of the 4th Int. Battelle Conf. on Remediation of
Chlorinated and Recalcitrant Compounds, May 24-27, Monterey, USA: 8 pages
SCHAD H., KLEIN R., ALT S., WEISS J., TIEHM A., MÜLLER A., SCHMITT-BIEGEL B. 2005.
Biosorption barrier at a former tar factory in Offenbach: (1) An innovative concept for
long-term in-situ treatment of highly contaminated groundwater.
In: Uhlmann O., Annokkée G., Arendt F. (Hrsg.) CONSOIL 2005, Proceedings (CD) of the
9th international FZK/TNO conference on soil-water systems, Bordeaux, 3-7 Oct. 2005:
1482-1486
S
TIEBER M., HAESELER F., WERNER P., FRIMMEL F. H. 1994.
A rapid screening method for micro-organisms degrading polycyclic aromatic
hydrocarbons in microplates.
Applied Microbiology Biotechnology 40: 753-755
T
IEHM A., SCHULZE S. 2003.
Intrinsic aromatic hydrocarbon biodegradation for groundwater remediation.
Oil & Gas Science and Technology – Rev. IFP 58 (4): 449-462
TIEHM A., STIEBER M., WERNER P., FRIMMEL F.H. 1997.
Surfactant-enhanced mobilization and biodegradation of polycyclic aromatic hydrocarbons
in manufactured gas plant soil.
Environmental Science & Technology 31, 9, 2570-2576.