...

GROSS WITTKE et al 2010 Temperature Effects on Bank Filtration

by user

on
Category: Documents
1

views

Report

Comments

Transcript

GROSS WITTKE et al 2010 Temperature Effects on Bank Filtration
Q IWA Publishing 2010 Journal of Water and Climate Change | 01.1 | 2010
55
Temperature effects on bank filtration: redox conditions
and physical-chemical parameters of pore water at Lake
Tegel, Berlin, Germany
A. Gross-Wittke, G. Gunkel and A. Hoffmann
ABSTRACT
In the city of Berlin, artificial groundwater recharge techniques, such as bank filtration and
infiltration ponds, are an important source for drinking water production. Climate change with
increasing surface water temperatures can influence the water purification processes during
bank filtration mainly due the intensification of metabolic processes leading to a decrease of
oxygen and an increase of anaerobic conditions. In Lake Tegel a significant increase of water
temperature in the epilimnion of 2.48C within the last 30 years was recorded. For a better
understanding of induced bank filtration at Lake Tegel, redox processes and physical-chemical
A. Gross-Wittke (corresponding author)
G. Gunkel
A. Hoffmann
Department of Water Quality Control,
Berlin University of Technology,
Sekr. KF 4, Straße des 17. Juni 135,
10623 Berlin,
Germany
Tel.: +49 30 31423791
Fax: +49314 30 23850
E-mail: [email protected]
conditions within the surface sediment layers (0–26 cm depth) at the littoral infiltration zone were
investigated. The influence of temperature in the range of 0 –258C on microbial catalysis of redox
processes, such as reduction of nitrate and sulphate, was examined during the period March
2004–June 2005. High water temperatures (16–258C) were accompanied by negative redox
potentials (EH ¼ 247 mV) and decreasing Ninorg concentrations, while the amount of ammonia,
Mn2 + and Fe2 + was rising. This indicates redox processes such as denitrification, Mn4 +
reduction, nitrate respiration and ammonification, as well as Fe3 + reduction. The reduction of
sulphate, however, has not yet become significant at Lake Tegel, but with increasing water
temperature, sulphate reduction must be expected.
Key words
| bank filtration, climate change, groundwater recharge, redox chemical processes,
self-purification, water temperature
INTRODUCTION
Bank filtration is an old water abstraction technology and
oxygenated for iron precipitation by water sprinkling, and
has been used for over 100 years. Different infiltration
fast sand filtration is used for iron elimination.
systems, such as river banks, lake shores and artificial ponds
Bank-filtration water purification is based on two
for groundwater recharge have been established (Gunkel &
processes: first, biological self-purification and second,
Hoffmann 2009). At present, riverbank filtration is the main
geochemical and hydrogeological processes during ground-
water abstraction method used in many European regions,
water passage (Gunkel & Hoffmann 2009). Biological self-
and lake bank filtration also is applied on a large scale in a
purification is realised by an adapted biocoenosis with high
few places such as Berlin, where 2.6 million inhabitants
turnover rates, whereas geochemical and hydrogeological
have been supplied with bank filtration water from Lake
processes are characterised by low turnover rates. However,
Tegel and other sites for about the past 70 years. Water
the transit time during bank filtration of about 50
abstraction occurs from over 1,000 wells. Only raw water is
days guarantees water purification by absorption and
doi: 10.2166/wcc.2010.005
A. Gross-Wittke et al. | Temperature effects on bank filtration
56
Journal of Water and Climate Change | 01.1 | 2010
metabolisation. The zone with increased bioactivity and its
regionally but show a clear trend toward warming (EEA
structural components within the interstices, the biofilm,
2007; IPCC 2008). For example, regions of the Iberian
amounts only to a few metres of the bank filtration zone
Peninsula and East Germany are already characterised by
(Brugger et al. 2001; Hiscock & Grischek 2002; Gunkel &
low precipitation and little excess of water (Schröter et al.
Hoffmann 2009). This observation is of great interest for
2005). The city of Berlin and the surrounding Brandenburg
water treatment by bank filtration, especially when con-
region is one of the driest regions in Germany, with a
sidering shock loads and physical and chemical disturb-
negative climate water balance in summer. Therefore, this
ances such as temperature increases.
area is highly sensitive to climate change (PIK et al. 2003).
The purification process within the pore water is a
In future, the mean discharge of the Havel and Spree
redox-dependent biodegradation of organic material, and
Rivers, with extended bank filtration sites, will further
turnover rates are highly dependent on local redox
decrease, and without this dilution effect, higher sewage
conditions as well as on temperature. Up to now, little has
and other contaminants will be increasingly concentrated
been discovered about these processes during groundwater
(Gunkel 2009), and will, thus, strongly influence water
recharge. An assessment of redox processes and tempera-
quality following bank filtration. Lake water warming was
ture effects for bank filtration is required for a better
registered at Lake Müggel, an enlargement of the Spree
understanding of the regulatory mechanisms of biological
River in Berlin, and within 25 years the mean summer
self-purification and physical-chemical redox processes, as
surface water temperature had risen about 2.38C (Adrian
well as their effects on pore water quality in natural and
et al. 2006), with consequences for oxygen and nutrient
induced bank filtration sites. Degradation of organic
levels (Wilhelm & Adrian 2008) and phytoplankton
material leads to consumption of oxygen, and the resultant
response (Huber et al. 2008).
use of alternative electron acceptors such as nitrate,
manganese, iron oxide, hydroxide and sulphate. This leads
to the formation of distinct redox zones along the flow
MATERIALS AND METHODS
direction (Massmann et al. 2008).
Temperature effects on bank filtration are highly
significant and have become the focus of further research
Regional climate
activities with special regard to climate change and water
The Berlin region is part of the temperate climate zone, with
temperature all over the world, as well as implementation of
transition to continental climate. The mean annual tem-
bank filtration techniques in some subtropical and tropical
perature is 8.98C, and the warmest months are July and
countries (Sens & Dalsasso 2007; Dash et al. 2008; Ray
August, with means of 18.58C and 17.78C, respectively
2008). Temperature change scenarios in Europe vary
(Table 1). Regular hot periods occur with air temperatures
Table 1
|
Annual mean climatic characteristics of the Brandenburg region, including Berlin
Summer meanp
Climatic feature
Winter mean†
Annual mean
Precipitation (mm a21)
314.4
262.1
603.5
Potential evaporation (mm a21)
517.8
110.2
628.0
Climate balance (mm a21)
2 176.4
224.5
151.9
Air temperature (8C)
14.7 (max ¼ 39.4)
2.7 (min ¼ 2 29.5)
8.7
18.5 (24.0 highest
monthly mean) (River Havel)
5.65
12.1
Sun duration (h d
21
)
Water temperature (8C)‡
p
4.2 – 4.7
May–October 1951– 2000.
November– April 1951–2000.
‡
Water temperature date from IfM (2009); all other data from PIK et al. (2003).
†
57
A. Gross-Wittke et al. | Temperature effects on bank filtration
up to 398C. Precipitation amounts to only 603 mm year21,
whereas actual evaporation is 511 mm year
21
.
Journal of Water and Climate Change | 01.1 | 2010
(reed Phragmites australis, water lily Nuphar lutea). On the
shore of Lake Tegel, water abstraction for Berlin’s water
It must be pointed out that occasionally a high epilimnic
supply is accomplished through several galleries with 116
water temperature occurred in the Berlin lakes: 28.98C
wells, 30 – 60 m deep, about 100 m from the lake. The
(Lake Tegel, this study) and .308C (Lake Müggel;
pumping rate of each well ranges from 50 to 150 m3 h21.
R. Adrian, pers. com.). A further temperature increase can
Experiments were conducted at the eastern shore of
lead to periodically ‘tropical’ conditions in this region.
Lake Tegel in front of the Reiswerder islet (N528340 1300 , E
Within a scenario until 2055 for Berlin-Brandenburg, the
138150 2500 to N528340 1000 , E138150 2400 ; Figure 1). The littoral
Potsdam Institute for Climate Impact Research estimates an
zone of Lake Tegel is formed by fine- to medium-sized
increasing number of hot days (.308C air temperature)
sands, which are poorly sorted. The theoretical sediment
with an increase in sun duration per day in summer
permeability values (kf) were calculated from sediment
compared with the reference period of 1951 –2000
grain-size distribution after Beyer (1964) and range from
(PIK et al. 2003).
5 £ 1025m s21 to 1 £ 1024 m s21. The smallest values are in
the upper sediment layer, owing to a higher proportion of
Study site
fine sands; kf values based on in situ infiltration rates and
hydraulic potential, measured in the upper 2 cm, were
Lake Tegel is a lowland lake, a glacial enlargement of the
4 £ 1027 to 6 £ 1025 m sec21, indicating severe plugging of
Havel River, situated in Berlin, with an area of 396 ha and a
the interstices (Table 2, Hoffmann & Gunkel 2009).
mean depth of 6.6 m. Water balance and water quality are
determined by the inflow of two small rivers (Tegeler Fließ,
Nordgraben) and by water exchange with the Havel River
Climate change effects in Lake Tegel
(Figure 1). After a severe eutrophication period, a phos-
Within 28 years (1980 –2008), the mean surface water
phate elimination plant was built in 1985, and now the
temperature (0.5 m water depth) of Lake Tegel had risen
water quality of the lake is mesotrophic, but organic-rich
about 2.48C (monthly mean). This climate change effect is
anoxic sediments are still typical, and periodically cyano-
obvious, with increasing minimum temperatures during
bacteria blooms occur (Heinzmann & Chorus 1994;
winter and increasing maximum temperatures in summer
Schauser et al. 2006). Lake Tegel has a sedge-rich littoral
(Figure 2). During the period studied, the maximum
zone with forest stands, sporadically interrupted by beach
epilimnic water temperatures reached about 23 –258C.
sections caused by erosion, with sparse macrophyte growth
Temperature at the infiltration site was not recorded
Figure 1
|
Map of Central Europe and the study area at Lake Tegel, showing the water depths in meters. Arrows mark water inflow and outflow (modified from Gunkel et al. 2009).
A. Gross-Wittke et al. | Temperature effects on bank filtration
58
Table 2
|
Journal of Water and Climate Change | 01.1 | 2010
Lake Tegel sediment characteristics of the littoral zone
Sediment depth (cm)
0–5
5–10
10 –15
15 –20
20 –25
25 –28
Water content (%)
20.3
20.2
11.5
14.5
17.6
16.8
Mean grain diameter (mm)
0.13
0.14
0.27
0.25
0.22
0.23
d10 (mm)
0.07
0.07
0.09
0.11
0.09
0.10
Sorting coefficient (So)
1.48
1.47
1.61
1.46
1.44
1.41
Porosity
0.40
0.40
0.26
0.31
0.36
0.35
kf (m sec21)p
5 £ 1025
5 £ 1025
7 £ 1025
1 £ 1024
8 £ 1025
1 £ 1024
p
According to Beyer (1964).
Source: Hoffmann & Gunkel (2009).
continuously, but we must assume that temperatures in the
For different sediment depths (at 2 – 18 cm), oxygen
shallow littoral zone are even higher than in the central lake
concentrations were determined via optodes: optical
monitoring station.
oxygen-minisensors (PST3, Precision Sensing GmbH),
connected with fibre optics extension cords, and an oxygen
meter (Fibox 3, Precision Sensing GmbH) were used. Seven
Field methods
optodes were permanently inserted horizontally into the
From March 2004 to June 2005, pore water sampling was
sediment fixed by an acrylic plate with holes drilled.
conducted monthly at a water depth of 30– 50 cm, about
Measurements were made monthly between November
3 m away from the splash water zone close to Phragmites
2004 and December 2005. Nevertheless the application of
and Nuphar stands. Pore water was recovered using micro-
the optodes in fine sediments needs further investigations
suction cups (diameter 2.5 mm, length 5 cm; UMS GmbH
due to high variability.
Munich, Germany). The suction cups were permanently
inserted in the interstices in 5 cm steps from 1 to 26 cm
Analytical methods
sediment depth, fixed in a rack-like acrylic glass tube. This
device was placed into the sediment, and the micro-suction
For the determination of pH and redox potential (EH) in
cups were inserted horizontally from inside to outside the
parallel with pore water sampling, a sediment core was
tube into the undisturbed sediment.
removed using an acrylic glass tube (diameter 6 cm), and
35.0
30.0
Monthly mean: T = 8.73 + 0.083 Year (n = 341, R 2 = 0.955)
Summer maximum: T = 20.7825 + 0.121155 Years (n = 28)
T (°C)
25.0
20.0
15.0
10.0
5.0
0.0
1980
1985
1990
1995
2000
2005
Years
Figure 2
|
Surface water heating in Lake Tegel, Berlin, since 1980; water depth 0.5 m, central lake position. Shown are the mean of the years (ANOVA variance analysis; p ¼ 0.09)
and the maximum summer temperatures (ANOVA variance analysis; p ¼ 0.01) (data from Berlin Senate Department for Urban Development).
59
A. Gross-Wittke et al. | Temperature effects on bank filtration
Journal of Water and Climate Change | 01.1 | 2010
redox potentials were measured by pushing the sediment in
the core sampler upwards step by step and inserting a WTW
pH and redox electrode (Pt in Ag/AgCl) every 2 cm into
the sediment.
Pore water and lake samples were filtered (0.45 mm pore
filter) and analysed immediately in the laboratory. Most
parameters were assayed by flow injection analysis (FIAstar
5000, Foss Tecator) because of small probe volumes of a few
þ
ml (Ntotal, NO2
3 , and NH4 ), and metal ion concentrations
were determined by atomic absorption spectroscopy (Varian Spektra A 400) after acidification with concentrated
nitric acid to pH , 2. Fe, Mn and Ca were analysed using
GBC 906A Scientific Equipment. SO22
4 was determined by
ionic chromatography (IC Dionex), and dissolved organic
carbon (DOC) was analysed using a High TOC Elementar
analyser. Nitrite analyses were carried out immediately after
sampling in situ using a test kit (Aquaquant, Merck). Lake
water was analysed in a similar way using the methods
mentioned above.
In analyses by scanning electron microscope with
energy dispersive spectroscopy (SEM-EDS), wet minicores (1 cm height, 0.7 cm diameter) with sediment were
air dried and then sputtered with gold or carbon. The minisediment cores were inserted completely into the vacuum
chamber of an REM-EDS Hitachi S 2700 electron microscope with an acceleration voltage of 20 kV and IDFix hardand software from SAMx for analysis.
RESULTS
Lake water
Lake Tegel is a weakly alkaline lake with a conductivity of
767.7 ^ 10.3 mS cm21, owing mainly to a high calcium
content of 70 mg L21 (Table 3). After treatment of inflow
water by a phosphate elimination plant, lake water quality
was enhanced and actually can be classified as mesotrophic,
but cyanobacterial blooms occur periodically (Schauser
et al. 2006). In lake water the main redox components
besides
O2
21
^ 0.67 mg L
(11.64 ^ 1.21 mg L21)
are
nitrate
(1.65
), manganese (17.40 ^ 13.72 mg L21), iron
(winter mean, 11.67 ^ 3.50 mg L21; summer mean, 77.33
^ 85.23 mg L21) and sulphate (116.80 ^ 14.08 mg L21).
Nitrate has been present for a long time in Lake Tegel
water and this points clearly to the eutrophication period,
with high Ninorg concentrations (up to 12 mg L21) but low
21
NO2
). The recuperation period
3 concentrations (,2 mg L
since 1985 has led to increasing NO2
3 concentrations (up to
6 mg L21), and since 1995 an oligotrophication period has
occurred with decreasing Ninorg concentrations and a lack
of ammonium, which means that Ninorg concentration
21
corresponds to a NO2
3 concentration of about 2.5 mg L
(Figure 3). During the summer period, NO2
3 concentration
decreases to 1.5 mg L21, and in some years there were no
occurrences or only traces of NO2
3.
Data statistical analyses were done by a nonlinear curve
fit as well as an extrapolation up to 308C water temperature
by OriginPro 7.5 software. A one-way ANOVA test
(OriginPro 7.5) was used for significant level determination.
Bank infiltration redox conditions and water
temperature
For further comparison between the electron avail-
The highest EH values were observed at temperatures from
ability and consumption within the infiltrating lake water
0 to 158C with maximum redox potentials in the range of
( ¼ input into the bank filtration system) and the pore water
þ348 mV to þ383 mV. The EH values decrease significantly
at . 20 cm sediment depth ( ¼ output into the ground-
with increasing temperatures in the total sediment zone
water), the redox-equivalent for each redox sensitive species
(0–10 cm); thus, immediately after infiltration pore-water
(O2; NO3; Mn/Fe; SO4) was calculated on the basis of
redox potential decreases, no depth gradient was detected
Equation (1):
in the upper infiltration zone (Figure 4). The redox
potentials for the temperature range of 16 – 208C varied
Redox equivalent ¼ m=M·z
p
ð1Þ
from þ178 to 214 mV, whereas at temperatures between
21 and 258C even lower EH values with þ17 mV and
m: concentration [mmol L
21
];
M: molar mass [mg mmol21];
z p: number of electrons transferred.
247 mV were observed.
The pore water concentrations of all investigated redoxactive compounds (except for SO22
4 ) exhibit temperature
A. Gross-Wittke et al. | Temperature effects on bank filtration
60
Table 3
|
Journal of Water and Climate Change | 01.1 | 2010
Lake Tegel water chemistry
Mean (2004/2005)
pH
Conductivity (mS cm21)p
Standard deviation
Min
Max
8.00
0.34
7.50
8.70
767.67
10.27
757.00
789.00
Ptot (mg L21)
0.06
0.04
0.03
0.20
Ca2 þ (mg L21)
70.38
11.55
48.40
80.00
Chla (mg L21)†
14.31
8.61
2.00
32.00
7.12
0.43
6.30
7.70
11.64
1.21
10.09
13.42
1.65
0.67
0.85
2.73
21
0.10
0.06
0.02
0.21
Ntot (mg L21)
1.77
0.66
0.98
2.92
DOC
21
O2 (mg L
)
NO3 (mg L21)
NH4 (mg L
2þ
Mn
(mg L
)
21
)
17.40
13.23
3.30
43.10
11.67
3.50
8.00
16.00
77.33
85.23
18.00
175.00
116.80
14.08
91.7
133.8
Fe2 þ (mg L21) winter
2þ
Fe
21
(mg L
) summer
SO4 (mg L21)
p
Data set May–September 2008 (LAGeSo 2008).
†
Data set May–September 2004 & 2005 (LAGeSo 2008).
dependencies, even if the effect for some of them is small
interstices (NO2
3 )
and for some compounds a high variance exists. The
depletion. The scattering of the data is important, because
influence of temperature on Ninorg and NO2
3 concentrations
a small-scale patchiness of sediment characteristics occurs
N2), which occurs after oxygen
in pore water and lake water is variable (Figure 5(a,b)): a
as well as a time-dependent variation of infiltration capacity
statistically significant decline of Ninorg and NO2
3 in
(Hoffmann & Gunkel 2009). The extrapolation of Ninorg and
the interstices from 0 to 58C in comparison with the
NO2
3 concentrations in pore water at 26– 308C indicates
temperature range 6– 258C was proven, whereas the lowest
a further decline, with increasing temperature down to
Ninorg and
NO2
3
21
values were detected at 11 –158C (0.62
0.8 mg L21 Ninorg, respectively NO2
3.
21
þ
The NO2
3 /NH4 redox process also was observed at a
2
NO2
3 ). The declining Ninorg and NO3 concentrations
Lake Tegel bank infiltration site: NHþ
4 concentration in
indicate a bacterial denitrification process within the
pore water rises with increasing water temperature while
^ 0.15 mg L
Ninorg,
respectively
0.59 ^ 0.15 mg L
14.0
Ninorg./ NO3– -N (mg L–1)
Ninorg (mg L–1)
12.0
NO3– -N (mg L–1)
10.0
8.0
6.0
4.0
2.0
0.0
1980
1985
1990
1995
2000
2005
Years
Figure 3
|
Ninorg and NO2
3 concentrations in surface water of Lake Tegel since 1980; water depth 0.5 m, central lake position (data from Berlin Senate Department for Urban
Development).
A. Gross-Wittke et al. | Temperature effects on bank filtration
61
Journal of Water and Climate Change | 01.1 | 2010
0.104 ^ 0.058 mg L21 Fe2 þ . However, for lake water and
pore water Fe2 þ concentrations an assimilable pattern was
observed. At temperatures up to 308C, the extrapolation of
manganese (0.449 mg L21 Mn2 þ ) and iron concentrations
(0.177 mg L21 Fe2 þ ) showed a further increase in Lake
Tegel pore water.
22
The SO22
redox reaction is still without signifi4 /S
cance in the pore water of the Lake Tegel bank filtration
site. The SO22
4 concentrations showed a slight increase with
rising temperature and did not differ from the lake water
(Figure 5(f)); the mean concentrations of SO22
4 varied from
21
116 ^ 5 mg L21 SO22
SO22
4 at 0 –58C to 130 ^ 9 mg L
4 at
21– 258C, possibly an increase owing to higher mineralisation rates in the summer period. A decrease of SO22
in
4
pore water interstices was not registered, but desulphurisation had already occurred in the interstices, proven by pyrite
crystallisation (Figure 6). At 5 – 6 cm depth, localized zones
Figure 4
|
Redox potentials (mV) at different temperatures within the sediment. Note
that the y-axis starts with negative values.
the lake water values do not exceed 0.21 mg L
21
NHþ
4
(Figure 6(c)). The NHþ
4 pore water concentrations with
with anaerobic conditions occurred, sulphate reduction was
enabled, indicating a limiting concentration of redox
components with a high EH potential, even while a high
infiltration flow velocity of 0.5 m d21 occurred.
values up to 0.77 mg L21 NHþ
4 at increased temperatures
(21 – 258C) are significantly (a ¼ 0.05) higher than at
0 – 158C. Thus a temperature effect occurs, indicating
NO2
3
bacterial ammonification as incomplete
respiration
in the pore water interstices. The extrapolation of the NHþ
4
concentration up to 308C shows a further increase with a
mean concentration of 0.44 mg L21 NHþ
4.
DISCUSSION
Climate change has led to an increase of lake water
temperature, and up to now water temperature has
increased to 2 – 38C during the past 50 years (EEA 2007).
Only few data are available about temperature increase of
The Mn2 þ and Fe2 þ concentrations demonstrate the
the epilimnic water, which indicates a more intensive
existence of further anoxic processes in the pore water
temperature increase, e.g. in Lake Müggel of 2.38C in 25
interstices such as redox-chemical reduction of Mn4 þ to
years (Adrian et al. 2006) and of 2.48C in Lake Tegel in 28
Mn2 þ as well as of Fe3 þ to Fe2 þ , particularly in the higher
years (this study). Wilhelm et al. (2006) investigation proved
temperature range of 21 –258C. At temperatures ,208C
that epilimnic lake water temperature is not simply
only traces of Mn2 þ and Fe2 þ occurred (, 0.05 mg L21),
correlated with air temperature but is a more complex
whereas at temperatures . 218C an intensive reduction of
multi-parameter interdependent with air temperature, air
4þ
Mn
and Fe
3þ
2þ
was observed (Figure 5(d,e)). The Mn
concentrations in the pore water as well as in lake water at
, 208C amounted to ,50 mg L
21
2þ
Mn
moisture and cloud cover.
Surface water warming must be regarded as a significant
, whereas at
factor in the bank filtration processes, because infiltration
21 –258C concentrations up to 311 mg L21 Mn2 þ in the
occurs only in shallow littoral zones (Massmann et al. 2008;
interstices occurred while the lake water concentrations
Gunkel & Hoffmann 2009); deeper (hypolimnic) lake areas
remains below 50 mg L21 Mn2 þ . For Fe2 þ pore water
are clogged by fine sediments. From an ecological point of
2þ
view, maximum epilimnic water temperature is regulating
for 16 – 208C were determined, a significantly higher
bank filtration processes, first by lethal effects of high tem-
21
concentrations with a mean of 0.030 ^ 0.011 mg L
concentration
(a ¼ 0.05)
at
21– 258C
occurred
Fe
with
peratures on organisms, and second by the intensification
62
A. Gross-Wittke et al. | Temperature effects on bank filtration
Journal of Water and Climate Change | 01.1 | 2010
(b) 3.0
(a) 3.0
y = A1*exp(–x/t1)+y0
y0 = 0.91;
A1 = 3.39;
t1 = 0.36
2.5
Inorg soluble N (mg L–1)
NO3 (mg L–1)
2.5
2.0
1.5
1.0
1.5
1.0
0.0
0.0
0–5
6–10
11–15 16–20 21–25
Temperature (°C)
0–5
26–30
(d)
0.8
y = A1*exp(–x/t1)+y0
y0 = 0.038;
A1 = 0.003;
t1 = 0.99
0.7
0.4
0.3
0.2
11–15 16–20 21–25
Temperature (°C)
26–30
y = A1*exp(–x/t1)+y0
y0 = 3.83;
A1 = 3.62;
t1 = 0.528
400
0.5
6–10
500
450
Dissolved Mn (µg L–1)
0.6
NH4 (mg L–1)
2.0
0.5
0.5
(c)
y = A1*exp(–x/t1)+y0
y0 = 1.00;
A1 = 3.62;
t1 = 0.37
350
300
250
200
150
100
0.1
50
0
0.0
0–5
6–10
11–15 16–20
Temperature (°C)
21–25
0–5
26–30
(e)
11–15 16–20 21–25
Temperature (°C)
26–30
(f)
200
160
y = A1*exp(–x/t1)+y0
y0 = 5.76;
A1 = 0.17;
t1 = 0.71
180
160
140
120
140
SO4 (mg L–1)
Dissolved Fe (µg L–1)
6–10
120
100
80
100
80
60
60
Pore water
In filtration water
Extra polation 30°C
40
40
20
20
0
0
0–5
Figure 5
|
6–10
11–15 16–20
Temperature (°C)
21–25
26–30
0–5
6–10
11–15 16–20 21–25
Temperature (°C)
26–30
Pore water ( ) and infiltration water ( ) concentration at different temperatures. Dashed line shows the nonlinear curve fit (equation box); white dot (W) gives
extrapolated concentration values for the temperature range 26 –308C; note that Mn2 þ and Fe2 þ concentrations are plotted in mg L21.
63
A. Gross-Wittke et al. | Temperature effects on bank filtration
Figure 6
|
Journal of Water and Climate Change | 01.1 | 2010
FeS2 formation in the interstices (5–6 cm depth) at Lake Tegel bank filtration site. (a): Beginning anaerobic conditions with distinct FeS2 crystals covering the extracellular
polymeric substances (EPS), which form net structures and plaques (right side, above). (b): Matured FeS2 crystals as framboidal pyrite formed in the EPS. SEM photos;
crystal determination by SEM-EDS.
of all metabolic processes in water, leading to a decrease of
oxygen and the probability of increased anaerobic conditions
in bank filtration zone. Anaerobic conditions cause lethal
damage to all aerobic microorganisms as well as to
meiofauna, the active components of self-purification processes in bank filtration. The anaerobic microbial community
is less effective, and metabolic processes are slower, and this
cannot substitute for aerobic bacteria and meiofauna activity.
3. SO2þ
4 reduction zone
1
1 þ 1
2
CH2 O þ SO22
4 þ H ! HS þ H2 O þ CO2
2
2
2
4. CH4 fermentation zone
1
1
CH2 O þ CO2 ! CH4 þ CO2
2
2
The main part of the microbial community lives within
Sediment redox chemical processes
the aerobic zone, and these organisms are primarily
responsible for the degradation of organic material. The
The theoretical model consists of an infiltrating lake water
anoxic zone is characterised by the intense use of
input flow towards the sediment, which contains oxygen,
alternative electron acceptors and processes like Fe3 þ ,
nitrate, manganese, iron and sulphate as well as fine
Mn4 þ , NO2
3 reduction as well as denitrification. The toxic
particulate organic matter and DOC. Distinct redox zones
SO24 þ reduction as well as CH4 fermentation zones with the
were defined along the flow direction, which proceed from
formation of H2S, S22, and CH4 cannot be tolerated by
the highest energy yield downwards. After sediment passage
the benthic community, and thus are unacceptable from
through these zones infiltrate/pore water finally reaches
the ecological point of view as well as for the drinking water
groundwater. Water quality of this output flow into
purification process.
groundwater strongly depends on the preceding redox and
Under undisturbed conditions, littoral lake sediments
degradation processes along the bank filtration (Figure 7).
possess relatively high redox potentials (approximately
The following major redox zones can be distinguished:
þ300 to þ400 mV) owing to low organic matter content
and resuspension of the sediment by wind and wave action.
1. Aerobic respiration zone
But at bank filtration sites, redox potential decreases by
CH2 O þ O2 ! CO2 þ H2 O
intensive infiltration of lake water with its nutrients and
2. Anoxic Fe3 þ , Mn4 þ , NO3 reduction zone
þ
CH2 O þ 8H þ 4FeðOHÞ3 ! 4Fe
2þ
þ 11H2 O þ CO2
CH2 O þ 2MnO2 þ 4Hþ ! 2Mn2þ þ 3H2 O þ CO2
4
4 þ
2
7
CH2 O þ NO2
3 þ H ! CO2 þ N2 þ H2 O
5
5
5
5
dissolved organic matter; the particulate organic matter
(POM), however, does not contribute to the turnover
processes, because the vertical transport is hindered by
the three-dimensional structure of the extracellular polymeric substances (EPS; Gunkel et al. 2009). Another factor
of significance for the decrease of the redox potential is the
water temperature: With increasing water temperature,
64
A. Gross-Wittke et al. | Temperature effects on bank filtration
|
Figure 7
Journal of Water and Climate Change | 01.1 | 2010
Theoretical model of biogenic redox processes in the lake infiltration site, modified after Andreae et al. (1989).
microbial activity increases, and the oxygen consumption
small and were located in dead-end pores of the interstices
within the sediment also increases. After oxygen depletion,
(Gunkel & Hoffmann 2009), but the often-discussed
alternative electron acceptors are used step by step (see
hypothesis (Gunkel et al. 2009) of organic-rich flocs with
Figure 7). Microbially catalysed reduction of nitrate and
a redox gradient from outside to inside could not be verified.
dentrification starts below EH values of þ421 mV, followed
Within the infiltrating lake water, the calculation of
by the reduction of manganese below EH values of
4þ
,
redox equivalents for e 2 acceptors like O2, NO2
3 , Mn
þ396 mV. With further decreasing redox potentials,
and Fe3 þ were 1.5 ^ 0.13 mmol L21 in total. During the
reduced iron appears, and finally desulphurisation starts at
sediment passage from 0 to 26 cm depth the available redox
EH values of 2 100 mV (Matthess 1990).
equivalents decreased to about 0.39 mmol L21. After the
This step by step occurrence of sediment-bound redox
depletion of this redox equivalent pool only the SO22
4 -redox
processes corresponds with the findings of this study at the
equivalents with a mean of 10.53 ^ 0.72 mmol L21 provides
bank filtration site of Lake Tegel. Especially at high
electrons for further reduction/degradation processes of
temperatures (21 – 258C), the pore water analysis revealed
organic matter, but under anaerobic conditions, the
4þ
denitrification, ammonification and the reduction of Mn
3þ
formation of toxic S22.
, but up to now the sulphate reduction has had no
From an ecological point of view, the sulphate reduction
significance; but sulphate reduction and occurrence of S22
is a catastrophe because of the extreme toxic effect of S22,
has already been demonstrated by pyrite crystallisation at a
and the interstitial biocoenosis with a high abundance
depth of 5– 6 cm. Possibly these anaerobic zones were very
and diversity will break down (Beulker & Gunkel 1996).
and Fe
65
A. Gross-Wittke et al. | Temperature effects on bank filtration
Journal of Water and Climate Change | 01.1 | 2010
The interstitial biocoenosis is the cause of the physical-
and its effects, or the role of nitrate as an electron acceptor
chemical structure of the interstices because of excretion of
in limnic systems.
EPS, which also accounts for self-purification processes
during bank filtration (Gunkel & Hoffmann 2009).
CONCLUSIONS
With regard to climate change, water warming can lead to
Temperature effects on bank filtration
The capacity of bank filtration is limited by the amount of
redox compounds with an EH . 0 mV, as they contain
mainly nitrate, manganese and iron. These redox processes
large-scale changes in natural self-purification capacities
during bank filtration. The consequences of increasing
water temperatures in bank filtration at Lake Tegel,
Germany, are:
are well known during lake eutrophication with an excess
1. a decrease in redox potential owing to intensive
of POM and the formation of anaerobic sediments in the
microbial and respiratory activity, and the development
hypolimnion of the lake. Additionally an increase in
of anoxic conditions;
temperature can lead to EH , 0 mV in sediments, because
2. the intense use of alternative electron acceptors after
of the reduced solubility of oxygen in water and the
oxygen consumption, such as NO2
3 microbial catalysed
intensification of biological as well as biochemical pro-
reduction of nitrate/ammonification, denitrification and
cesses. Temperature dependency of redox processes such as
reduction of Mn4 þ and Fe3 þ followed by enrichment of
denitrification was reported in various studies (e.g. Carrera
groundwater with manganese, and iron;
et al. 2003); the maximum denitrification rates were
continuously increasing with rising temperatures.
However, with further increase in maximum summer
temperature a reduction of
SO22
4
and the formation of H2S
3. no significant sulphate reduction so far, but the occurrence of S22 was already shown by pyrite crystallisation
within the sediment.
In future, with further increase in maximum summer
must be expected. Studies of bacterial sulphate reduction
temperature the expected SO4 reduction could lead to the
indicate that environmental temperature is the dominant
breakdown of the interstitial biocoenosis, and by this way of
variable that influences the sulphate reduction rate in
the self-purification capacity.
sediment; furthermore, optimum temperatures of 30 –378C
are reported for sulphate reduction (Abdollahi & Nedwell
1979; Ingvorsen et al. 1981).
The future situation of the Lake Tegel bank filtration
site depends, on the one hand, on low concentration of
nitrate in water mainly from geological conditions in the
watershed and by an additional continuous reduction of
nitrate due to the implementation of high-efficiency
wastewater treatment plants, and on the other hand by
dramatically increased temperatures in the epilimnic water.
At times complete depletion of nitrate in the water occurs,
ACKNOWLEDGEMENTS
We thank the German Scientific Society (DFG) for
financing this project (Gu 307/11-1/2). The investigations
were supported by Dipl. Ing. Ulrich Gernert, ZELMI and
TU Berlin (SEM-EDS). The Senate of Berlin performed the
long-term lake monitoring and provided these data, and
Cand.-Ing. Florian Selge carried out statistical work on
Lake Tegel water-quality data.
and other redox compounds such as manganese and iron
served in the infiltration zone as electron acceptors, but the
capacity of electron acceptors under anoxic conditions is
nearly depleted (to 26% of the input) and the occurrence of
anaerobic conditions must be expected in future. Recent
risk assessment and management strategies are not sufficiently developed concerning water temperature change
REFERENCES
Abdollahi, H. & Nedwell, D. B. 1979 Seasonal temperature as a
factor influencing bacterial sulphate reduction in a saltmarsh
sediment. Microb. Ecol. 5, 73 –79.
Adrian, R., Wilhelm, S. & Gerten, D. 2006 Life-history traits of lake
plankton species may govern their phenological response to
climate warming. Glob. Environ. Change 12, 625– 661.
66
A. Gross-Wittke et al. | Temperature effects on bank filtration
Andreae, M. O., Schimel, D. S. & Robertson, G. P. 1989 Report of
the Dahlem Workshop on Exchange of Trace Gases between
Terrestrial Ecosystems and the Atmosphere, Berlin, February
19 –24 1989.
Beulker, C. & Gunkel, G. 1996 Studies on the occurrence and
ecology of meiofauna and benthic algae in the littoral
sediment of lakes. Limnologica 26, 311 –326.
Beyer, W. 1964 Determination of water permeability of gravels and
sands from the grain size distribution curve (Zur Bestimmung
der Wasserdurchlässigkeit von Kiesen und Sanden aus der
Kornverteilungskurve.). WWT (Wasserwirtschaft und
Wassertechnik) 14 (6), 165 –168.
Brugger, A., Reitner, B., Kolar, I., Quéric, N. & Herndl, G. J. 2001
Seasonal and spatial distribution of dissolved and particulate
organic carbon and bacteria in the bank of an impounding
reservoir on the Enns River, Austria. Freshw. Biol. 46 (8),
997– 1016.
Carrera, J., Vicent, T. & Lafuente, F. J. 2003 Influence of
temperature on dentrification of an industrial high-strength
nitrogen wastewater in a two-sludge system. Water SA 29
(1), 11– 16.
Dash, R. R., Mehrotra, I., Kumar, P. & Grischek, T. 2008 Lake bank
filtration at Nainital, India: water quality evaluation.
Hydrogeol. J. 16, 1089 –1099.
EEA (European Environment Agency) 2007 Climate change and
water adaptation issues. Techn. Report 2. European
Environment Agency, Copenhagen.
Gunkel, G. 2009 Vulnerability of aquatic systems to climate chance—
findings from European waters. In: Sustainable Development:
Energy, Environment and Natural Disasters (eds L. Duarte &
P. Pinto), pp. 9–23. Fundação Luis de Molina, Évora.
Gunkel, G. & Hoffmann, A. 2009 Bank filtration of rivers and lakes
to improve the raw water quality for drinking water supply.
In Water Purification (ed. N. Gertsen & L. Sønderby),
pp. 137 –169. Nova Science, Hauppauge, NY.
Gunkel, G., Beulker, C., Hoffmann, A. & Kosmol, J. 2009 Fine
particulate organic matter (FPOM) transport and processing in
littoral interstices—use of fluorescent markers. Limnologica 39,
185– 199.
Heinzmann, B. & Chorus, I. 1994 Restoration concept for Lake
Tegel, a major drinking and bathing water resource in a
densely populated area. Environ. Sci. Technol. 28, 1410 –1416.
Hiscock, K. M. & Grischek, T. 2002 Attenuation of groundwater
pollution by bank filtration. J. Hydrol. 266, 139– 144.
Hoffmann, A. & Gunkel, G. 2009 Bank filtration in the sandy
littoral zone of Lake Tegel (Berlin): Structure and dynamics of
the biological active filter zone and clogging processes.
Limnologica (doi:10.1016/j.limno.2009.12.003).
Huber, V., Adrian, R. & Gerten, D. 2008 Phytoplankton response to
climate warming modified by trophic state. Limnol. Oceanogr.
53, 1–13.
Journal of Water and Climate Change | 01.1 | 2010
IfM (Institut für Meterologie der Freien Universität Berlin) 2009
Meteorological mean and extreme values of Berlin-Dahlem—
Watertemperature of River Havel (Meterologische Mittel- und
Extremwerte von Berlin-Dahlem—Wassertemperatur der
Havel). http://www.met.fu-berlin.de/~manfred/extrema.htm
#wasser
Ingvorsen, K., Zeikus, J. G. & Brock, T. D. 1981 Dynamics of
bacterial sufhate reduction in an eutrophic lake. Appl.
Environ. Microbiol. 42 (6), 1029 –1036.
IPCC (Intergovernmental Panel on Climate Change) 2008 Climate
change and Water. IPCC Technical Paper VI. UNEP.
LAGeSo (Berlin State Office of Health and Socials) 2008 Data set
bathing water monitoring Lake Tegel 1994 –2008.
Massmann, G., Nogeitzig, A., Taute, T. & Pekdeger, A. 2008 Seasonal
and spatial distribution of redox zones during lake bank filtration
in Berlin, Germany. Environ. Geol. 54 (1), 53–65.
Matthess, G. 1990 The Composition of Groundwater. 2nd edition (Die
Beschaffenheit des Grundwassers. 2. Auflage), Lehrbuch der
Hydrologie 2, S. 498, Gebrüder Borntraeger, Berlin, Stuttgart.
PIK (Potsdam Institute for Climate Impact Research),
Gerstengarbe, F. W., Badeck, F., Hattermann, F., Krysanova,
V., Lahmer, W., Lasch, P., Stock, M., Suckow, F., Wechsung,
F. & Werner, P. C. 2003 Study of the climatic development in
Brandenburg until 2055, and the impacts on water balance,
agriculture and forestry, as well as first perspectives (Studie
zur klimatischen Entwicklung im Land Brandenburg bis 2055
und deren Auswirkungen auf den Wasserhaushalt, die Forstund Landwirtschaft sowie die Ableitung erster Perspektiven).
PIK Report 83, S.91.
Ray, C. 2008 Worldwide potential of riverbank filtration. Clean
Technol. Environ. Policy 10 (3), 223 –225.
Schauser, I., Chorus, I. & Heinzmann, B. 2006 Strategy and current
status of combating eutrophication in two Berlin lakes for
safeguarding drinking water resources. Water Sci. Technol. 54
(11 – 12), 93 –100.
Schröter, D., Acosta-Michlik, L., Arnell, A. W., Araújo, M. B.,
Badeck, F., Bakker, M. Bondeau, A., Bugmann, H., Carter, T.
& de la Vega-Leinert, A. C. 2005 ATEAM Final Detailed
report, related to overall project duration. Potsdam Institute for
Climate Impact Research. Potsdam.
Sens, M. L. & Dalsasso, R. L. 2007 Bank filtration of reservoir as an
efficient technique for water supply treatment. In Reservoir
and River Basin Management: Exchange of Experiences from
Brazil, Portugal and Germany (ed. G. Gunkel & M. C.
Sobral), pp. 253– 264. Universitätsverlag TU Berlin.
Wilhelm, S. & Adrian, R. 2008 Impact of summer warming on the
thermal characteristics of a polymictic lake: consequences for
oxygen, nutrients and phytoplankton. Freshw. Biol. 53, 226–237.
Wilhelm, S., Hintze, T., Livingstone, D. M. & Adrian, R. 2006 Longterm response of daily epilimnetic temperature extrema to
climate forcing. Can. J. Fish. Aquat. Sci. 63 (11), 2467 –2477.
First received 14 May 2009; accepted in revised form 11 February 2010
Fly UP