- Email: [email protected]
Emissions of porewater compounds and gases from the subaquatic sediment disposal site “Rodewischhafen”, Hamburg harbour
~ Pergamon
Wat. Sci. Tech. Vol. 37, No. 6-7, pp. 87-93,1998. ~
PIT: S0273-1223(98)OO186-3
J998 IAWQ. Published by Elsevier Science LId Printed in Great Britain. 0273-1223198 SI9'00 + 0-00
EMISSIONS OFPOREWATER COMPOUNDS AND GASES FROM THE SUBAQUATIC SEDIMENT DISPOSAL SITE "RODEWISCHHAFEN", HAMBURG HARBOUR M. Kussmaul*, A. Groengroeft* and H. Koethe**
* Institute ofSoil Science, University of Hamburg, Allende Platz 2,20146 Hamburg, Germany ** Federallnstitute ofHydrology, Box 309, 56003 Koblenz. Germany ABSTRACT In the year 1993 a confined and unused harbour basin was used to store 290,000 m 3 of fine-grained dredged material from Hamburg harbour. About 70% of the deposit surface was water covered. The edge areas were above the water table and covered with reed. Emissions of dissolved compounds into the groundwater, as well as surface gas emissions were measured from 1994 to 1996. As indicators for water fluxes from the deposit we used NH4+ and HCOf because of their high concentrations in mud porewater in comparison to groundwater. The average concentrations of NH4+ and HCOf in the porewater increased during 2 years from 85 to 250 mg NH4+ I-I and from 2.0 to 3.1 g HCOf I-I, while the groundwater samples showed constant values of 8 mg NH 4+ I-I and 0.7 g HCOf I-I. Furthermore. the average gas emissions over the water surface were 3.2 g CH4 m- 2 d- I and 0.8 g C02 m- 2 d-I. In contrast, no methane and 3.0 g CO2 m- 2 d- I were emilled from land areas. The results indicated. that there were no significant emissions of mud porewater compounds into the groundwater but high CH4-emissions over the water covered surface of the mud deposit. © 1998 IAWQ. Published by Elsevier Science Ltd
KEYWORDS
Contaminated sediment; greenhouse gas emission; porewater emission; subaquatic disposal. INTRODUCTION A confined and unused harbour basin with a surface area of 42,000 m2 was used to store 290,000 m3 of fine• grained dredged material from Hamburg harbour below the water table. The harbour mud is polluted mainly by heavy metals and arsenic and is characterised by high concentrations of organic compounds, between 3 to 7% TOC. The decay of carbon leads to anaerobic conditions and thus to gas production in the mud deposit. Existing gas inclusions as well as a high content of clay and silt (50 - 98%) result in a low water permeability of the mud. The aim of the investigation was to find out if the mud deposit contributes under these storage conditions to the groundwater pollution. Additionally, production- and emissionrates of the greenhouse gases methane and carbondioxide were determined. High CH 4-emissions can cause explosive 87
88
M. KUSSMAUL et al.
°
gasmixtures (5-15% CH 4 in at least 11.6% 2, Rettenberger 1994) in confined areas. This could be the case, if the deposit area would be used for construction purposes. METHODS Site description Figure I shows Ii sketch of the harbour sediment deposit ,Rodewischhafen" with surrounding sand- and low permeable estuarine clay layers. The mud was stored mainly subaquatic, but about 30% of the area had a land surface which was covered with reed. The average depth of the deposit was 7 m below the water table. Throughout the year the water table changed about ± 5 cm. The monitoring well RM2, which was located 10 m from the deposit rim, was used to collect groundwater samples from different depths. The resulting direction of the tidal influenced groundwater is shown in Figure 1. The groundwater layer II has a broad distribution and is influenced by the tide, whereas layer I is only locally distributed and shows almost stagnant conditions. At the down-gradient position RM2 a conventional monitoring well with 20 m screen length (10-30 m below water table) was used to collect groundwater samples from layer II by pumping. In layer I a multi-tube well with 3 monitoring points (3-5 m below water table) was installed, the samples were collected by a vacuum pump device. reed
•
depth [m] 4
sand, ~ clay, ~ harbour mud
Figure I. Sketch of the subaquatic harbour mud deposit "Rodewischhafen" with sampling site RM2-1 (groundwater layer I) and RM2-11 (groundwater layer II). The deposit contained 290,000 m 3 mud with a surface area of 42,ooom 2. Samp!jn~
techniques
Porewater samples. Porewater samples (50 ml) were collected with suction Iysimeters, permanently installed at depths of 3, 4 and 5 m below water table. To avoid precipitation of cations or phosphates, the collection bottles were prefilled with I ml HN0 3 (32%) or 2 ml H2S04 (38%) respectively. For other anions and ammonium no prefilling was carried out. The chemical parameters were analysed afterwards in the laboratory . Chemical and physical analysis Measurement of ground- and porewater compounds. The chemical composition of the water samples were determined by using standard analytical procedures: titration (HCOf), anionchromatography (Cl-, S042-, N0 3-), photometry (NH 4+, P042-), AES (Na+, K+) and AAS (Mg2+, Ca2+, Fe2+, Mn 2+, Zn 2+, Cd 2+, Cu 2+, Cr 3+, Ni 2+, Pb 2+, As).
Porewater compounds and gases emission
89
Atmospheric pressure and temperature. The atmospheric pressure was measured with a barometer GDH 12A (Greisinger electronic). A thermometer type GTH 100/2 with Ni-CrNi antenna (Greisinger electronic) was used to determine water and mud temperatures.
Gas analysis The concentrations of methane, carbondioxide and oxygen were measured with a gaschromatograph type Fisons GC 6000. A Fill-detector was used to determine low concentrations of methane « 0.5%). Higher CH4-eoncentrations as well as 02 and CO 2 were detected by a WLD-detector. The variation of the measurement was 2% for CH 4 and CO2 and 5% for 02' The detection parameters were: Columns: Carrier gasflow: Temperatures:
Haye Sep 0, 20 ft, mesh 100/120 and Molsieb SA, 7 ft, mesh 60/80 32 ml helium min- l Oven 70°C, injector 100 C. WLD 100°C filaments 180°C. Fill
=
=
=
=
=200°C.
Gas production and ~as emission Gas production. Mud cores were taken with a beeker sampler out of the centre of the deposit in 0 to 4 m depth. A gas tight 130 ml glasbottle with rubber septum was filled with 80 ml of mud from different depths and immediately flushed with nitrogen. The increase of CH 4 and C02 in the headspace was measured after 12 h of incubation for the following 48 h. During that time the samples were incubated at the corresponding depth temperature of the deposit in the dark. Afterwards the bottle was refilled with water. The weight of the added water corresponds to the gasvolume in the bottle. The gas productionrates were calculated by a linear regression in [g m· 3 fw mud d· l ]. Gas emission. Gas boxes were used to determine the gas emissions over water- and land surface (Figure 2). The increase of CH4 and CO 2 in the box was measured for 6 h. Every 45 min a gasvessel was replaced and afterwards analysed in the laboratory. The gas emissionrates were calculated by linear regression in [g m- 3 fw mud dol].
Figure 2. Gas boxes to determine gas emissions over the water and land surface.
RESULTS AND DISCUSSION Chemjcal parameters of the ~round- and porewater Table 1 shows the porewater composition as median values over time (Oct. 1994 to Sept. 1996) and sample depth (1-5 m below deposit surface). Due to the high TOC of the harbour mud and the absence of oxygen, nitrate and sulphate have been totally reduced. The concentration of chloride is still in the range of the water from the River Elbe which was primarily used for flushing the dredged material in the harbour basin.
90
M. KUSSMAUL et al.
Table I. Average porewater concentrations of anions. cations and trace metals from 1994 to 1996 anions [mg r1]
cr
sol-
N03-
pol-
HC03" measured HC03calculated n.d.
cations [mg r 1]
heavy metals and As [Ilg r 1] Zn2+ 16.5 Cd2+ <0.1 Cu2+ 1.2 Cr+ 4.7 Ni2+ 4.5
40 n.d. n.d. 1.4 1810
NH.+ Na+ K+ M 2+
Ca~+
189 107 35 71 428
2470
Fe2+
64
Pb 2+
Mn2+
16
As
=not detectable.
<3.0 6.2
The main anionic constituent of the pore water is bicarbonate (90% of the sum of anions). Part of the HCOf is lost due to the sampling technique (precipitation in the Iysimeter tubes as Ca- or Fe-carbonates. gaseous loss as CO 2), therefore the concentration of HCOf was calculated from ion balance. The concentrations of ammonium were high. single values varying between 66 and 300 mg I-I. Phosphate behaves irregular. partly not being detectable. partly showing considerable concentrations up to 10 mg I-I. Potassium. magnesium and calcium have been markedly increased in comparison to the River Elbe water whereas the content of sodium is still in the range of riverine water. The low redox potentials of the sediments results in increased concentrations of mobile iron and manganese. The concentrations of heavy metals are low. only the solubility of arsenic has to be considered. Vertical differences in the pore water composition were rather small. Due to the mineralisation of organic material and subsequent geochemical processes (carbonate buffering. cation exchange) the concentrations of HCOf. NH4+. Ca2+. Na+. K+ and Mg2+ have been significantly (P < 0.05) increased since the beginning of the investigation. From regression analysis on all individual values (n 32 to 40) the mean annual increase was calculated for HCOf with 9.40. NH4+: 4.66. Na+: 0.50. K+: 0.15. Mg2+: 0.37 and Ca2+: 1.47 (values in mmol I-I a-I). These calculated trends were nearly charge balanced (anions + 9.40 mmol eq I-I a-I. cations + 9.00 mmol eq I-I a-I).
=
From the measured change in porewater composition it was possible to calculate the potential release of gaseous C from mineralisation on a unit area. Using a water content of 70% fresh weight. a particle density of t 2.55 g cm- 3• a depth of the disposed mud of h = 6 m and an organic carbon content of TOC = 4.1 % the unit area consists of 2.2 t m- 2 dry matter and 5.14 m3 m- 2 pore water with the dry matter containing 90 kg m• 2 organic carbon. Using two different C/N-ratios (C/N-ratio measured = 10.5 mol mol-I. C/N-ratio according to Redfield 6.625 mol mol-I) the measured increase in ammonium content in pore water (4.66 mmol I-I a• I) results in a net-C-mineralisation of 3.0 kg C m-2 a-lor 1.9 kg C m- 2 a-I. Part of the mineralised carbon leads to an increase in alkalinity (9.4 mmol I-I a-I). thus the potential release of gaseous C has to be reduced to 2.44 kg C m- 2 a-lor 1.33 kg C m- 2 a-I. These calculations showed. that the daily rates of C-mineralisation may vary between 3.7 and 6.7 g C m- 2 d-I. It is yet not considered. which part of the mineralised N is adsorbed as NH4+-cation by matrix (leading to an increase in calculated C-mineralisation) and to which content the increase in alkalinity is caused by carbonate buffering thus not decreasing the mineralised C.
=
=
The influence of the subaquatic disposal on adjacent groundwater layers should be easily detectable by a change in bicarbonate or ammonium content. As shown above these parameters are highly concentrated in the sediment pore water whereas the background values of typical Elbe river concentrations are low (HC0 3-: 338 mg I-I. NH4+: 8.7 mg I-I. Groengroeft and Miehlich. 1995). Figure 3 shows the concentrations of these indicator parameters in groundwater layer I and II. The concentrations vary slightly with time but are still in the low range of bank infiltrated water. In layer I the quality is worse than in layer II. as illustrated by arsenic. but the compositions of anions and cations indicate. that there is no direct response on the mud
Porewater compounds and gases emission
91
disp?sal. Thus it can be concluded that the influence of subaquatic disposal on groundwater quality is lackmg or rather small. This result corresponds with the low permeability of the disposed mud and the observed termination of the consolidation process (Schwieger pers. comm.). 300 250 ::"
.:..
~ +
2"
200 150 100 50
~--
10 5
..... 0 !,2OOO ~1600
~
1200
0'" BOO
~ .coo
...~...
...
...
...
...
...
...
...
...
...
...
...
0 12
.....
.:..
~ ..(
9 6
3
0
...
1994 94/95 - . - porewater - 0 - groundwater layer I
1995
95/96
1996
- & . - groundwater
layer II
Figure 3. NH4+-. measured HCOf- and As-concentrations in the ground- and porewater from 199410 1996. The dala of porewaler and groundwater layer I are the mean value of 3 parallel measurements whereas each data point of layer II represents one single measuremenl.
Gas productjon and
~as
emissjon
Gas productionrates of the mud deposit from different depths are shown in Figure 4. Methane production varied between 0 and 2 g CH4 m- 3 fw mud d-I. No CH4 was formed in the upper oxic layer of the deposit. However, the CH4-production corresponds to the course of the temperature profile. In contrast, the CO 2• production ranged between 2 and II g CO 2 m- 3 fw mud d- I. The lowest COrProduction occurred in the upper deposit layer. The anaerobic decay of carbon in deeper mud layers leads obviously to a higher CO 2• formation. The release of methane and carbondioxide into the atmosphere over the water- and land surface is shown in Figure 5. Extremely high CH4-emissions between I and 7 g m- 2 d- I were measured over the water surface. The highest CH4-emission was observed in summer. The winter emissions were reduced to I g CH 4 m- 2 d- I because of the low mud temperatures which declined to an average winter minimum of 9°C (data not shown). In summer, the methane emissions occurred mainly in form of gas bubbles coming out of the mud. Therefore, the resting time of methane in oxic mud- and water layers was too short for an effective reduction of methane by methanotrophic bacteria. No significant CH 4-emissions could be observed over the land surface. There was obviously a very effective methane oxidising bacterial community in the mud layer
92
M. KUSSMAUL et al.
above the water table, which was able to oxidise the methane coming from deeper layers. In comparison, natural Elbe marshland shows CH 4-emissionrates ofO-O.OS g CH 4 m- 2 d- I (Pfeiffer, 1994).
~C02 ~CH
4
- - temperature
o
2
4
6
8
gas production [g m- 3 fw mud d-I]
10
12
temperature [0C]
Figure 4. Gas productionrates of the fresh mud with the corresponding depth temperatures from the 24. April 1996.
7 6
IwaterI
- . - air pressure
~C02 ~CH4
5
1035 1030 1025
4 3
1020 1015
25 20 4
15
3
10
2
5
°t--.....---,...----r---,--..,.---"T-----iO July. Aug.
Oct.
Dec. 1995 I
Apr.
June
1996
Figure 5. Gas emissions over the water- and the land surface with the corresponding air temperatures and atmospheric pressures.
Porewater compounds and gases emission
93
The emissions of CO 2 over the water surface showed with 0-1.2 g m- 2 d-I only little variation throughout the year (Figure 5). In contrast, the land surface emissions of 1.5-4.5 g CO 2 m- 2 d- I were higher. However, the land CO 2-emissions were in the range of natural marshland emissions of 1.1-5.5 g CO 2 m- 2 d- I (Nyman and DeLaune, 1991). Only a small part of the produced CO 2 (see Figure 4) was emitted over the surface areas. Reasons are the good solubility of CO 2 in water and its chemical reactions with ions and surfaces. Additionally, CO2 is fixed by autotrophic bacteria and plants for growth. CONCLUSIONS The subaquatic deposition of dredged material offers a good opportunity to store contaminated harbour sediments. No emissions of contaminants into the groundwater could be observed in this case, mainly because of the low water permeability of the mud. The concept of subaquatic disposal, as proposed by Bertsch & Knoepp (1990), can be supported with our data. Also the high emissions of the greenhouse gases methane and carbondioxide can be reduced by an oxic soil layer which covers the deposit above the water table. The use of a plant cover promotes aeration of the upper soil and binds CO 2, REFERENCES Bertsch, W. and Knoepp. H. (1990). Ober die Unterbringung von Baggergut in Kiesgruben als "Unterwasser-deponie". Z. dt. geol. Ges. 141, 393-398. Groengroeft, A. and Miehlich. G. (1995). F1iichenhafter Stoffeintrag durch aufgehllhte Gebiete im Bereich der Hamburger Elbmarsch. Z. dt. geoL Ges. 146,235-242. Nyman, J. A. and DeLaune, R. D. (1991). C02 emission and soil Eh responses to different hydrological conditions in fresh. brackish, and saline marsh soils. Limnol. Oceanogr. 36(7).1406-1414. Pfeiffer, E.-M. (1994). Methane fluxes in natural wetlands (marsh and moor) in Northern Gennany. Current Topics Wetland Biogeochem. I, 35-46. Rettenberger. G. (1994). Sicherheitstechnische Erfordernisse an Deponieentgasungsanlagen aufgrund der GUY 17.4. In: Trierer Berichte zur Abfallwirtschaft, Eifassung und Nutzung von Deponiegas, G. Rettenberger and R. Stegmann (eds.), vol 6. Economica Verlag Bonn. Gennany. pp. 223-237. Schwieger, F. (1996). Personal communication. Bundesanstalt fUr Wasserbau Hamburg, Gennany.
Wat. Sci. Tech. Vol. 37, No. 6-7, pp. 87-93,1998. ~
PIT: S0273-1223(98)OO186-3
J998 IAWQ. Published by Elsevier Science LId Printed in Great Britain. 0273-1223198 SI9'00 + 0-00
EMISSIONS OFPOREWATER COMPOUNDS AND GASES FROM THE SUBAQUATIC SEDIMENT DISPOSAL SITE "RODEWISCHHAFEN", HAMBURG HARBOUR M. Kussmaul*, A. Groengroeft* and H. Koethe**
* Institute ofSoil Science, University of Hamburg, Allende Platz 2,20146 Hamburg, Germany ** Federallnstitute ofHydrology, Box 309, 56003 Koblenz. Germany ABSTRACT In the year 1993 a confined and unused harbour basin was used to store 290,000 m 3 of fine-grained dredged material from Hamburg harbour. About 70% of the deposit surface was water covered. The edge areas were above the water table and covered with reed. Emissions of dissolved compounds into the groundwater, as well as surface gas emissions were measured from 1994 to 1996. As indicators for water fluxes from the deposit we used NH4+ and HCOf because of their high concentrations in mud porewater in comparison to groundwater. The average concentrations of NH4+ and HCOf in the porewater increased during 2 years from 85 to 250 mg NH4+ I-I and from 2.0 to 3.1 g HCOf I-I, while the groundwater samples showed constant values of 8 mg NH 4+ I-I and 0.7 g HCOf I-I. Furthermore. the average gas emissions over the water surface were 3.2 g CH4 m- 2 d- I and 0.8 g C02 m- 2 d-I. In contrast, no methane and 3.0 g CO2 m- 2 d- I were emilled from land areas. The results indicated. that there were no significant emissions of mud porewater compounds into the groundwater but high CH4-emissions over the water covered surface of the mud deposit. © 1998 IAWQ. Published by Elsevier Science Ltd
KEYWORDS
Contaminated sediment; greenhouse gas emission; porewater emission; subaquatic disposal. INTRODUCTION A confined and unused harbour basin with a surface area of 42,000 m2 was used to store 290,000 m3 of fine• grained dredged material from Hamburg harbour below the water table. The harbour mud is polluted mainly by heavy metals and arsenic and is characterised by high concentrations of organic compounds, between 3 to 7% TOC. The decay of carbon leads to anaerobic conditions and thus to gas production in the mud deposit. Existing gas inclusions as well as a high content of clay and silt (50 - 98%) result in a low water permeability of the mud. The aim of the investigation was to find out if the mud deposit contributes under these storage conditions to the groundwater pollution. Additionally, production- and emissionrates of the greenhouse gases methane and carbondioxide were determined. High CH 4-emissions can cause explosive 87
88
M. KUSSMAUL et al.
°
gasmixtures (5-15% CH 4 in at least 11.6% 2, Rettenberger 1994) in confined areas. This could be the case, if the deposit area would be used for construction purposes. METHODS Site description Figure I shows Ii sketch of the harbour sediment deposit ,Rodewischhafen" with surrounding sand- and low permeable estuarine clay layers. The mud was stored mainly subaquatic, but about 30% of the area had a land surface which was covered with reed. The average depth of the deposit was 7 m below the water table. Throughout the year the water table changed about ± 5 cm. The monitoring well RM2, which was located 10 m from the deposit rim, was used to collect groundwater samples from different depths. The resulting direction of the tidal influenced groundwater is shown in Figure 1. The groundwater layer II has a broad distribution and is influenced by the tide, whereas layer I is only locally distributed and shows almost stagnant conditions. At the down-gradient position RM2 a conventional monitoring well with 20 m screen length (10-30 m below water table) was used to collect groundwater samples from layer II by pumping. In layer I a multi-tube well with 3 monitoring points (3-5 m below water table) was installed, the samples were collected by a vacuum pump device. reed
•
depth [m] 4
sand, ~ clay, ~ harbour mud
Figure I. Sketch of the subaquatic harbour mud deposit "Rodewischhafen" with sampling site RM2-1 (groundwater layer I) and RM2-11 (groundwater layer II). The deposit contained 290,000 m 3 mud with a surface area of 42,ooom 2. Samp!jn~
techniques
Porewater samples. Porewater samples (50 ml) were collected with suction Iysimeters, permanently installed at depths of 3, 4 and 5 m below water table. To avoid precipitation of cations or phosphates, the collection bottles were prefilled with I ml HN0 3 (32%) or 2 ml H2S04 (38%) respectively. For other anions and ammonium no prefilling was carried out. The chemical parameters were analysed afterwards in the laboratory . Chemical and physical analysis Measurement of ground- and porewater compounds. The chemical composition of the water samples were determined by using standard analytical procedures: titration (HCOf), anionchromatography (Cl-, S042-, N0 3-), photometry (NH 4+, P042-), AES (Na+, K+) and AAS (Mg2+, Ca2+, Fe2+, Mn 2+, Zn 2+, Cd 2+, Cu 2+, Cr 3+, Ni 2+, Pb 2+, As).
Porewater compounds and gases emission
89
Atmospheric pressure and temperature. The atmospheric pressure was measured with a barometer GDH 12A (Greisinger electronic). A thermometer type GTH 100/2 with Ni-CrNi antenna (Greisinger electronic) was used to determine water and mud temperatures.
Gas analysis The concentrations of methane, carbondioxide and oxygen were measured with a gaschromatograph type Fisons GC 6000. A Fill-detector was used to determine low concentrations of methane « 0.5%). Higher CH4-eoncentrations as well as 02 and CO 2 were detected by a WLD-detector. The variation of the measurement was 2% for CH 4 and CO2 and 5% for 02' The detection parameters were: Columns: Carrier gasflow: Temperatures:
Haye Sep 0, 20 ft, mesh 100/120 and Molsieb SA, 7 ft, mesh 60/80 32 ml helium min- l Oven 70°C, injector 100 C. WLD 100°C filaments 180°C. Fill
=
=
=
=
=200°C.
Gas production and ~as emission Gas production. Mud cores were taken with a beeker sampler out of the centre of the deposit in 0 to 4 m depth. A gas tight 130 ml glasbottle with rubber septum was filled with 80 ml of mud from different depths and immediately flushed with nitrogen. The increase of CH 4 and C02 in the headspace was measured after 12 h of incubation for the following 48 h. During that time the samples were incubated at the corresponding depth temperature of the deposit in the dark. Afterwards the bottle was refilled with water. The weight of the added water corresponds to the gasvolume in the bottle. The gas productionrates were calculated by a linear regression in [g m· 3 fw mud d· l ]. Gas emission. Gas boxes were used to determine the gas emissions over water- and land surface (Figure 2). The increase of CH4 and CO 2 in the box was measured for 6 h. Every 45 min a gasvessel was replaced and afterwards analysed in the laboratory. The gas emissionrates were calculated by linear regression in [g m- 3 fw mud dol].
Figure 2. Gas boxes to determine gas emissions over the water and land surface.
RESULTS AND DISCUSSION Chemjcal parameters of the ~round- and porewater Table 1 shows the porewater composition as median values over time (Oct. 1994 to Sept. 1996) and sample depth (1-5 m below deposit surface). Due to the high TOC of the harbour mud and the absence of oxygen, nitrate and sulphate have been totally reduced. The concentration of chloride is still in the range of the water from the River Elbe which was primarily used for flushing the dredged material in the harbour basin.
90
M. KUSSMAUL et al.
Table I. Average porewater concentrations of anions. cations and trace metals from 1994 to 1996 anions [mg r1]
cr
sol-
N03-
pol-
HC03" measured HC03calculated n.d.
cations [mg r 1]
heavy metals and As [Ilg r 1] Zn2+ 16.5 Cd2+ <0.1 Cu2+ 1.2 Cr+ 4.7 Ni2+ 4.5
40 n.d. n.d. 1.4 1810
NH.+ Na+ K+ M 2+
Ca~+
189 107 35 71 428
2470
Fe2+
64
Pb 2+
Mn2+
16
As
=not detectable.
<3.0 6.2
The main anionic constituent of the pore water is bicarbonate (90% of the sum of anions). Part of the HCOf is lost due to the sampling technique (precipitation in the Iysimeter tubes as Ca- or Fe-carbonates. gaseous loss as CO 2), therefore the concentration of HCOf was calculated from ion balance. The concentrations of ammonium were high. single values varying between 66 and 300 mg I-I. Phosphate behaves irregular. partly not being detectable. partly showing considerable concentrations up to 10 mg I-I. Potassium. magnesium and calcium have been markedly increased in comparison to the River Elbe water whereas the content of sodium is still in the range of riverine water. The low redox potentials of the sediments results in increased concentrations of mobile iron and manganese. The concentrations of heavy metals are low. only the solubility of arsenic has to be considered. Vertical differences in the pore water composition were rather small. Due to the mineralisation of organic material and subsequent geochemical processes (carbonate buffering. cation exchange) the concentrations of HCOf. NH4+. Ca2+. Na+. K+ and Mg2+ have been significantly (P < 0.05) increased since the beginning of the investigation. From regression analysis on all individual values (n 32 to 40) the mean annual increase was calculated for HCOf with 9.40. NH4+: 4.66. Na+: 0.50. K+: 0.15. Mg2+: 0.37 and Ca2+: 1.47 (values in mmol I-I a-I). These calculated trends were nearly charge balanced (anions + 9.40 mmol eq I-I a-I. cations + 9.00 mmol eq I-I a-I).
=
From the measured change in porewater composition it was possible to calculate the potential release of gaseous C from mineralisation on a unit area. Using a water content of 70% fresh weight. a particle density of t 2.55 g cm- 3• a depth of the disposed mud of h = 6 m and an organic carbon content of TOC = 4.1 % the unit area consists of 2.2 t m- 2 dry matter and 5.14 m3 m- 2 pore water with the dry matter containing 90 kg m• 2 organic carbon. Using two different C/N-ratios (C/N-ratio measured = 10.5 mol mol-I. C/N-ratio according to Redfield 6.625 mol mol-I) the measured increase in ammonium content in pore water (4.66 mmol I-I a• I) results in a net-C-mineralisation of 3.0 kg C m-2 a-lor 1.9 kg C m- 2 a-I. Part of the mineralised carbon leads to an increase in alkalinity (9.4 mmol I-I a-I). thus the potential release of gaseous C has to be reduced to 2.44 kg C m- 2 a-lor 1.33 kg C m- 2 a-I. These calculations showed. that the daily rates of C-mineralisation may vary between 3.7 and 6.7 g C m- 2 d-I. It is yet not considered. which part of the mineralised N is adsorbed as NH4+-cation by matrix (leading to an increase in calculated C-mineralisation) and to which content the increase in alkalinity is caused by carbonate buffering thus not decreasing the mineralised C.
=
=
The influence of the subaquatic disposal on adjacent groundwater layers should be easily detectable by a change in bicarbonate or ammonium content. As shown above these parameters are highly concentrated in the sediment pore water whereas the background values of typical Elbe river concentrations are low (HC0 3-: 338 mg I-I. NH4+: 8.7 mg I-I. Groengroeft and Miehlich. 1995). Figure 3 shows the concentrations of these indicator parameters in groundwater layer I and II. The concentrations vary slightly with time but are still in the low range of bank infiltrated water. In layer I the quality is worse than in layer II. as illustrated by arsenic. but the compositions of anions and cations indicate. that there is no direct response on the mud
Porewater compounds and gases emission
91
disp?sal. Thus it can be concluded that the influence of subaquatic disposal on groundwater quality is lackmg or rather small. This result corresponds with the low permeability of the disposed mud and the observed termination of the consolidation process (Schwieger pers. comm.). 300 250 ::"
.:..
~ +
2"
200 150 100 50
~--
10 5
..... 0 !,2OOO ~1600
~
1200
0'" BOO
~ .coo
...~...
...
...
...
...
...
...
...
...
...
...
...
0 12
.....
.:..
~ ..(
9 6
3
0
...
1994 94/95 - . - porewater - 0 - groundwater layer I
1995
95/96
1996
- & . - groundwater
layer II
Figure 3. NH4+-. measured HCOf- and As-concentrations in the ground- and porewater from 199410 1996. The dala of porewaler and groundwater layer I are the mean value of 3 parallel measurements whereas each data point of layer II represents one single measuremenl.
Gas productjon and
~as
emissjon
Gas productionrates of the mud deposit from different depths are shown in Figure 4. Methane production varied between 0 and 2 g CH4 m- 3 fw mud d-I. No CH4 was formed in the upper oxic layer of the deposit. However, the CH4-production corresponds to the course of the temperature profile. In contrast, the CO 2• production ranged between 2 and II g CO 2 m- 3 fw mud d- I. The lowest COrProduction occurred in the upper deposit layer. The anaerobic decay of carbon in deeper mud layers leads obviously to a higher CO 2• formation. The release of methane and carbondioxide into the atmosphere over the water- and land surface is shown in Figure 5. Extremely high CH4-emissions between I and 7 g m- 2 d- I were measured over the water surface. The highest CH4-emission was observed in summer. The winter emissions were reduced to I g CH 4 m- 2 d- I because of the low mud temperatures which declined to an average winter minimum of 9°C (data not shown). In summer, the methane emissions occurred mainly in form of gas bubbles coming out of the mud. Therefore, the resting time of methane in oxic mud- and water layers was too short for an effective reduction of methane by methanotrophic bacteria. No significant CH 4-emissions could be observed over the land surface. There was obviously a very effective methane oxidising bacterial community in the mud layer
92
M. KUSSMAUL et al.
above the water table, which was able to oxidise the methane coming from deeper layers. In comparison, natural Elbe marshland shows CH 4-emissionrates ofO-O.OS g CH 4 m- 2 d- I (Pfeiffer, 1994).
~C02 ~CH
4
- - temperature
o
2
4
6
8
gas production [g m- 3 fw mud d-I]
10
12
temperature [0C]
Figure 4. Gas productionrates of the fresh mud with the corresponding depth temperatures from the 24. April 1996.
7 6
IwaterI
- . - air pressure
~C02 ~CH4
5
1035 1030 1025
4 3
1020 1015
25 20 4
15
3
10
2
5
°t--.....---,...----r---,--..,.---"T-----iO July. Aug.
Oct.
Dec. 1995 I
Apr.
June
1996
Figure 5. Gas emissions over the water- and the land surface with the corresponding air temperatures and atmospheric pressures.
Porewater compounds and gases emission
93
The emissions of CO 2 over the water surface showed with 0-1.2 g m- 2 d-I only little variation throughout the year (Figure 5). In contrast, the land surface emissions of 1.5-4.5 g CO 2 m- 2 d- I were higher. However, the land CO 2-emissions were in the range of natural marshland emissions of 1.1-5.5 g CO 2 m- 2 d- I (Nyman and DeLaune, 1991). Only a small part of the produced CO 2 (see Figure 4) was emitted over the surface areas. Reasons are the good solubility of CO 2 in water and its chemical reactions with ions and surfaces. Additionally, CO2 is fixed by autotrophic bacteria and plants for growth. CONCLUSIONS The subaquatic deposition of dredged material offers a good opportunity to store contaminated harbour sediments. No emissions of contaminants into the groundwater could be observed in this case, mainly because of the low water permeability of the mud. The concept of subaquatic disposal, as proposed by Bertsch & Knoepp (1990), can be supported with our data. Also the high emissions of the greenhouse gases methane and carbondioxide can be reduced by an oxic soil layer which covers the deposit above the water table. The use of a plant cover promotes aeration of the upper soil and binds CO 2, REFERENCES Bertsch, W. and Knoepp. H. (1990). Ober die Unterbringung von Baggergut in Kiesgruben als "Unterwasser-deponie". Z. dt. geol. Ges. 141, 393-398. Groengroeft, A. and Miehlich. G. (1995). F1iichenhafter Stoffeintrag durch aufgehllhte Gebiete im Bereich der Hamburger Elbmarsch. Z. dt. geoL Ges. 146,235-242. Nyman, J. A. and DeLaune, R. D. (1991). C02 emission and soil Eh responses to different hydrological conditions in fresh. brackish, and saline marsh soils. Limnol. Oceanogr. 36(7).1406-1414. Pfeiffer, E.-M. (1994). Methane fluxes in natural wetlands (marsh and moor) in Northern Gennany. Current Topics Wetland Biogeochem. I, 35-46. Rettenberger. G. (1994). Sicherheitstechnische Erfordernisse an Deponieentgasungsanlagen aufgrund der GUY 17.4. In: Trierer Berichte zur Abfallwirtschaft, Eifassung und Nutzung von Deponiegas, G. Rettenberger and R. Stegmann (eds.), vol 6. Economica Verlag Bonn. Gennany. pp. 223-237. Schwieger, F. (1996). Personal communication. Bundesanstalt fUr Wasserbau Hamburg, Gennany.