Abstract

Treatment of sludge in sludge treatment reed bed systems includes dewatering and mineralization. The mineralization process, which is driven by microorganisms, produces different gas species as by-products. The pore space composition of the gas species provides useful information on the biological processes occurring in the sludge residue. In this study, we measured the change in composition of gas species in the pore space at different depth levels in vertical sludge residue profiles during a resting period of 32 days. The gas composition of the pore space in the sludge residue changed during the resting period. As the resting period proceeded, atmospheric air re-entered the pore space at all depth levels. The methane (CH4) concentration was at its highest during the first part of the resting period, and then declined as the sludge residue became more dewatered and thereby aerated. In the pore space, the concentration of CH4 often exceeded the concentration of carbon dioxide (CO2). However, the total emission of CO2 from the surface of the sludge residue exceeded the total emission of CH4, suggesting that CO2 was mainly produced in the layer of newly applied sludge and/or that CO2 was emitted from the sludge residue more readily compared to CH4.

INTRODUCTION

Sludge treatment reed bed (STRB) systems are holistic sludge treatment facilities, which combine dewatering and mineralization. A major part of the experience on STRBs comes from Denmark (Nielsen & Willoughby 2005), which today houses more than 100 STRB facilities. However, other countries, e.g. Sweden (Gustavsson & Engwall 2012), Spain (Uggetti et al. 2009), Italy (Masciandaro et al. 2015), England (Nielsen & Cooper 2011), France (Vincent et al. 2011), Poland (Obarska-Pempkowiak et al. 2003), have also implemented STRBs.

An STRB consists of a number of separate beds, commonly eight or 10, but this number, and the area of the beds, can vary according to demand for treatment capacity, sludge quality and climatic conditions (Nielsen & Willoughby 2005; Uggetti et al. 2010). The beds are planted with common reed (Phragmites australis (Cav.) Trin. Ex Steud). The beds are loaded with sludge in turns: one bed is loaded with sludge for a number of days (‘loading period’), whereafter the loading shifts to another bed. When all beds have been loaded, the first bed in the row is loaded again; this loading cycle is continually repeated. The period between two loading periods of a bed is called a ‘resting period’.

Each bed in an STRB can be loaded with sludge for several years before it must be emptied. The solid part of the sludge accumulates in the bed as layers of sludge residues. After loading a bed, a large part of the water contained by the sludge drains off through the layers of sludge residue within a few hours (Nielsen 2003). The reject water is collected by reject water pipes embedded in a filter in the bottom of the beds. The reject water pipes also lead oxygen (O2) to the lower part of the sludge residue layer. During the subsequent resting period, the water content of the sludge residue declines further due to evapotranspiration (Nielsen 1990). The stems of the reeds make cracks in the surface of the sludge residue, through which water can evaporate and O2 enter the upper part of the sludge residue layer (Nielsen 2003).

Mineralization of the organic part of the sludge residue by microorganisms is an important part of the treatment process. This biological activity leads to a production of different gas species: Aerobic conditions in the sludge residue enhance production of carbon dioxide (CO2), while anaerobic conditions enhance production of methane (CH4). Nitrous oxide (N2O) is mostly related to denitrifications under anaerobic conditions, but can also be produced under aerobic conditions (Lloyd et al. 1987; Robertson et al. 1995), or in the transition phase between aerobic and anaerobic conditions (Yu et al. 2010).

The composition of gas species in air retrieved from the pore space in sludge residue could be informative in relation to understanding the dynamics of microbial processes during the resting period in the loading cycle. The aim of the present study was to measure the composition of gas species in the pore space at different depth levels in situ in sludge residue in an STRB. Gas species to be measured in the pore space were O2, CH4, CO2 and N2. For this purpose, gas probes made of stainless steel were inserted into the sludge residue at different depth levels and samples of pore space air extracted. To be able to relate the gas composition in pore space air at different depth levels to the total emission of gas from the surface of the sludge residue, static surface flux chambers were used to measure surface emissions of CO2, CH4 and N2O.

MATERIAL AND METHODS

Description of the experimental site

All measurements were undertaken at the STRB at Helsinge wastewater treatment plant (WWTP) in Denmark (56°01′15N; 12°19′49E). The daily operation of the STRB has been supervised by Orbicon A/S since 1996. Originally, the system was built with 10 beds, but in 2013, four new beds were added. Until 2010, the system was loaded with a mixture of two sludge types, namely surplus activated sludge (SAS) (sludge produced from biological treatment of domestic wastewaters) produced by Helsinge WWTP, and SAS produced by four to seven external WWTPs. The externally produced SAS was stored in concentration tanks at the specific WWTPs and was eventually transported to a common storage tank at Helsinge WWTP. Due to the storage and transportation process, SAS produced at the external WWTPs was more concentrated and anaerobic compared to SAS produced at Helsinge WWTP. The sludge mixture (consisting of SAS from Helsinge WWTP and sludge from other smaller WWTPs) had a dry solid content of 0.6–0.8%. However, since 2010, Helsinge STRB has mainly been loaded with SAS from Helsinge WWTP; occasionally SAS from the smaller WWTPs was mixed in. The change of sludge type in 2010 means that the deepest layers (formed before 2010) of sludge residue in the beds originate from the mixed sludge type, while the uppermost layers (formed after 2010) primarily originate from SAS produced at Helsinge WWTP.

When sludge is loaded onto the sludge residue in a bed, the main part of the water contained by the sludge drains off within 5 to 10 hours after the loading. Typically, the draining rate achieves a maximum of 0.01–0.02 L·s−1 m2 during dewatering of a batch. During a loading period, in which a bed receives sludge every day, the maximum draining rate declines. For the bed chosen for our study, the average maximum drain rate over a loading period is 0.01 L·s−1 m2. Table 1 provides details on system characteristics of Helsinge STRB.

Table 1

System characteristics of Helsinge STRB

System characteristics Quality (feed sludge) 
Population equivalent (PE) 42,000 Sludge type SAS 
Number of beds 14 DS (%) 0.6–0.8 
Capacity (ton DS y−1630 Loss on ignition (%) 45–65 
Total bed area (m214,700 Fat (mg kg−1 DS) 500–7,000 
Individual bed area (m21,050 Oil (mg kg−1 DS) 100–2,000 
Loading rate – dim. (kg DS m−2 y−160 Sludge age (aerobic days) 20–25 
Loading rate – real (kg DS m−2y−146–83 Start of operation period 1996 
Loading/resting (days) 4–7/44–75 Quality (sludge residue) 
Sludge batches per day in loading period 1–2 DS (mg kg−1 DS) 22 
Sludge batch size (m3150 Loss on ignition (%) 40–50 
Emptying periods 1) 2005–2008 Fat (mg kg−1 DS) 100 
2) 2011, 2013–2017 Oil (mg kg−1 DS) 100–500 
System characteristics Quality (feed sludge) 
Population equivalent (PE) 42,000 Sludge type SAS 
Number of beds 14 DS (%) 0.6–0.8 
Capacity (ton DS y−1630 Loss on ignition (%) 45–65 
Total bed area (m214,700 Fat (mg kg−1 DS) 500–7,000 
Individual bed area (m21,050 Oil (mg kg−1 DS) 100–2,000 
Loading rate – dim. (kg DS m−2 y−160 Sludge age (aerobic days) 20–25 
Loading rate – real (kg DS m−2y−146–83 Start of operation period 1996 
Loading/resting (days) 4–7/44–75 Quality (sludge residue) 
Sludge batches per day in loading period 1–2 DS (mg kg−1 DS) 22 
Sludge batch size (m3150 Loss on ignition (%) 40–50 
Emptying periods 1) 2005–2008 Fat (mg kg−1 DS) 100 
2) 2011, 2013–2017 Oil (mg kg−1 DS) 100–500 

Bed no. 5 was chosen as the experimental site. During the experimental period, this bed contained 80 cm of sludge residue. The experimental period ran from September 14th to October 15th, 2015, being the first and the last day of the resting period, respectively. Commonly, the resting periods at Helsinge STRB have a duration of 44–75 days. However, due to practical reasons, the resting period chosen as the experimental period was shortened to 32 days. Four measuring stations were installed in the bed (Figure 1(a)). At each of the four stations, a nest of gas probes and a flux chamber were installed. To minimize disturbance of the sludge residue and the surrounding reeds, wooden boardwalks were used to access the measuring stations.

Figure 1

(a) Bed no. 5 at Helsinge STRB system. The crosses indicate the location of the four measurement stations. (b) A schematic cross-section of the sludge residue and underlying drainage filter layer at a measuring station. At each station, a static surface flux chamber and a nest of five gas probes, covering different depths, were installed.

Figure 1

(a) Bed no. 5 at Helsinge STRB system. The crosses indicate the location of the four measurement stations. (b) A schematic cross-section of the sludge residue and underlying drainage filter layer at a measuring station. At each station, a static surface flux chamber and a nest of five gas probes, covering different depths, were installed.

Measurements of gas composition in pore space

At each measuring station, five gas probes (Ø 7 mm) were installed on the first day of the resting period (September 14th 2015) and stayed in the bed until the end of the resting period. At that time, the bed contained 100 cm of sludge residue. However, the sludge residue found 80–100 cm below the surface originates from sludge loaded into the bed in 1996–1998, while the sludge residue found 10–80 cm below the surface originates from sludge loaded into the bed in 2007–2014. Due to the time gap between these two sections of sludge, they are not comparable. Therefore, we excluded the sludge residue found 80–100 cm below the surface from the study.

To measure gas composition at different depths, the five gas probes at each measuring station were pushed down into the sludge residue to five different depth levels: 10, 20, 40, 60 and 80 cm below the surface (Figure 1(b)). The ends of the probes inserted into the sludge residue had three narrow slits, through which air from the pore space could enter. The first measuring date was scheduled as the second day of the resting period. Afterwards, measurements were conducted at day 8, 11, 15, 18, 22, 29 and 32 of the resting period. Measurements were undertaken at one measuring station at a time. On a measuring date, the measurements at the first station were initiated around 8–9 am and the measurements at the fourth station finished around 10–11 am.

Air samples from the pore space were collected using a syringe attached to the top of the probe using a 5 cm PVC plastic tube (Ø 7 mm). Before a sample of pore space air was drawn from a probe, the probe was emptied for stagnant air by use of the syringe. Afterwards, the syringe was flushed with atmospheric air and a sample of pore space air (30–100 ml) drawn from the probe. When shifting to a new probe, the plastic tube and the syringe were flushed with atmospheric air. Air samples were immediately transferred to a portable biogas analyzer (BIOGAS 5000). This analyzer measures the percentage composition of CO2, CH4, O2 and nitrous gas species. The nitrous gas species are mainly constituted of free nitrogen (N2), but also contain N2O and ammonia (NH3). However, N2 constitutes the major part; the nitrous gas species are therefore referred to as ‘N2’.

Measurement and data processing of surface gas emissions

At each measurement station, a static surface flux chamber was installed (Figure 1(b)) close to the nest of gas probes. The flux chambers were designed as described in Uggetti et al. (2012): the bottoms were cut out of four opaque plastic barrels (Ø 38 cm, H 40–45 cm) (Figure 2(b)). The barrels had tightly fitted lids, each equipped with a battery-driven fan (10 cm Ø) to ensure mixing of gas in the headspace, as suggested by Maljanen et al. 2007. Each lid was equipped with two sampling ports – one to aid in a mobile flux-measuring device and one for a thermometer.

Figure 2

Gas concentration profiles of methane, carbon dioxide, oxygen and residual gas species (mainly dinitrogen) at 10, 20, 40, 60 and 80 cm depths of the sludge residue in Bed no. 5 at Helsinge STRB system. Gas concentration profiles are shown for day 2, 8, 11, 15, 18, 22, 29 and 32 of a resting period. Gas concentrations are given as mean values of four measuring stations. The bars represent standard errors.

Figure 2

Gas concentration profiles of methane, carbon dioxide, oxygen and residual gas species (mainly dinitrogen) at 10, 20, 40, 60 and 80 cm depths of the sludge residue in Bed no. 5 at Helsinge STRB system. Gas concentration profiles are shown for day 2, 8, 11, 15, 18, 22, 29 and 32 of a resting period. Gas concentrations are given as mean values of four measuring stations. The bars represent standard errors.

Like the gas probes, the chambers were installed on the first day of the resting period, and stayed in the bed until the end of the resting period. To avoid gas leakage, the edges of the chambers were pushed approximately 10 cm into the sludge residue. When the chambers were mounted in the sludge residue, the reeds growing inside the chambers were cut to a height of 40–45 cm, making it possible to close the chambers with tightly fitting lids. Grünfeld & Brix (1999) showed that cutting of reeds did not affect gas emission rates significantly. The lids were only put on the chambers during measuring sessions (approximately 10 minutes); at all other times they were left open.

Surface gas emissions were measured from the flux chambers on the same dates as described for the gas probes. On a measuring day, the measurement of the flux chamber at the first measuring station was initiated at 8–9 am and the measurement at the fourth spot finished at 10–11 am.

When measuring surface gas emissions in a flux chamber, the fan in the lid was activated and the lid attached to the chamber approximately two minutes before initiation of the measurement. Concentrations of CO2, CH4 and N2O in the chamber were then recorded every minute during a measuring period of 5–10 minutes, using a mobile photoacoustic gas monitor (Gas Monitor INNOVA 1312 (Innova AirTech Instruments). The temperature in each flux chamber was noted and used when converting the measured gas concentrations from ppb to mol.

For all measuring days, gas emission rates (mol·m−2·d−1) of CO2, CH4 and N2O were calculated for each chamber. The rates were calculated as the linear increase in gas concentration over time. Regression lines with r2 < 0.8 were excluded. For each measuring day, a daily, average emission rate for each gas species was estimated by calculating the mean emission rate of the four flux chambers.

RESULTS

Figure 2 shows the pore space gas concentration profiles of N2, O2, CO2 and CH4, measured in the sludge residue in Bed no. 5 at STRB. The concentrations are averages of the concentrations measured at the four stations.

Concentration of free nitrogen in depth profiles

When a bed is loaded with sludge, the surface of the sludge residue is covered by a layer of new sludge. During the first days after application, the newly applied sludge has a muddy texture and could act as a ‘seal’ through which atmospheric air cannot enter the sludge residue, as suggested by Olsson et al. (2014). As the newly applied sludge becomes more dewatered, it develops a more porous texture, allowing atmospheric air to enter (Dominiak et al. 2011). The concentration of N2 in the pore space reflects the supply of atmospheric air. Free nitrogen is also produced in the sludge residue because of denitrification; however, this contribution will be minor in comparison to atmospheric N2. At days 2, 8 and 11 of the resting period, N2 constituted about 20–30% of the pore space air collected from 10 cm depth (Figure 2). From day 15 until day 32, the concentration increased, being almost 80% at day 32. This shows that air enters the sludge residue from the atmosphere. At 80 cm depth, the N2 content of the gas samples collected on days 2, 8, 11, 15, 18 and 22 constituted approximately 45–60% of the pore space air, while it increased to approximately 80% on days 29 and 32. The sludge residue at 80 cm depth is closer to the reject water pipes, through which atmospheric air is led to the lowest part of the sludge residue. The results thus confirmed that this part of the sludge residue was well aerated. During the first part of a resting period, dewatering of the sludge residue is mainly due to drainage. As long as reject water is traveling through the sludge residue, the lowest part of the sludge residue will receive water, which restrict atmospheric air from entering the pore space in this area. Later in the resting period, the water left in the sludge residue is mainly pore water and the drainage process fades out. However, the sludge residue is still undergoing dewatering, but now mainly due to evapotranspiration (Nielsen 1990). Our results support this, as the percentage of N2 increased in the last two days of the resting period; at this time no more reject water travels through the lowest layer of sludge residue and more air is allowed to enter.

The lowest content of N2 was found at 20 cm depth on day 8 and 11. This could be due to the ‘sealing effect’ created by the layer of newly applied sludge. At 40 cm depth, N2 always constituted 40–50% of the pore air. At 40 cm, the distance to both the surface and the reject water pipes was at its maximum. A likely explanation could be that the air in the pore space at this depth was supplied mainly by the rhizomes of the reeds rather than by a combination of rhizomes and diffusion through the sludge residue matrix; therefore, it was less affected by loadings and infiltrating water.

Concentration of oxygen in depth profiles

Oxygen in the sludge residue originates from the atmosphere and is supplied together with N2. Indeed, the fluctuations of the content of O2 found in the different depth levels at different dates resembles the fluctuations described for N2 (Figure 2): The O2 content measured at 10 cm depth was at its lowest on day 2 (8%) and increased as the resting period proceeded, constituting 17% of the pore air on day 32. On days 2 to 22, the content of O2 at 80 cm depth fluctuated between 10 and 15%, except for day 11, where the content only constituted 7% of the pore space air. On days 10 to 32, the content of O2 measured at 40 cm depth was the lowest observed among the different depth levels.

Concentration of carbon dioxide and methane in depth profiles

The sources of CO2 and CH4 in sludge residue are microbial metabolic processes. Until day 15, gas concentrations indicated that the major part of CH4 was produced at 10 cm depth (Figure 2). After day 15, the CH4 concentration at this depth decreased. From this time and onward, the major part of the CH4 was seen at 20 or 40 cm depth. During the resting period, it was observed that the amount of CH4 measured at 10 cm depth decreased, while the amount of O2 increased. Methane is produced under anaerobic conditions, meaning that this gas species becomes less dominant as atmospheric air, and thereby O2, enters the layer of newly applied sludge residue. This is due to less production of methane as anaerobic zones or pockets diminish but is also a result of dilution with atmospheric air. For the entire resting period, the amount of CH4 at 80 cm depth was low compared to the other depth levels. Again, this was consistent with the relatively high concentration of O2 measured at this depth. From day 18, the highest concentrations of CH4 were seen at 40 or 60 cm depth, also being the depth with the lowest concentrations of O2.

Carbon dioxide is produced by aerobic respiration. However, the metabolic pathway producing CH4 (methanogenesis) also produces CO2 in the ratio of about 3:2 (Jeris & McCarty 1965). It seems that the fluctuations of CO2 concentrations observed at the different dates and different depth levels followed the same pattern as CH4, the concentrations of CH4 in general just being higher (Figure 2), suggesting that CO2 was mainly produced by methanogenesis, rather than aerobic respiration. The higher concentrations of CH4 in comparison to CO2 could also be caused by removal of CO2 by dissolution in infiltration water. Other studies have also shown that CH4 tends to be trapped in the substrate where it is produced and is therefore released through pulses rather than as a steady flow (Chanton & Whiting 1995).

Surface emissions of methane, carbon dioxide and nitrous oxide

The surface emissions of CH4 and CO2 measured using static flux chambers (Figure 3) revealed that except for day 11, the surface emissions of CO2 always exceeded that of CH4. An explanation for this could be that a part of the CO2 surface emission detected by the flux chambers was produced in the layer of newly-applied sludge. The gas composition in this layer was not covered by the probe measurements, as this layer was too thin for obtaining a gas sample without withdrawal of atmospheric air. Fresh sludge contains large amounts of readily biodegradable carbon (C). After application, the newly applied sludge is in close contact with the atmosphere, which leads to intensive activity by aerobic microorganisms generating CO2. Also, CH4 generated in the sludge residue could be oxidized to CO2 in aerobic zones and in the upper 10 cm of the residue, which would result in higher CO2 surface fluxes in comparison to CH4 surface fluxes.

Figure 3

Surface emissions of carbon dioxide, methane and nitrous oxide from Bed no. 5 at Helsinge STRB system. The x-axis represents the amount of days passed since initiation of a resting period. Emissions are given as mean values of four measuring stations. The bars represent standard errors. Notice different scales on the y-axes.

Figure 3

Surface emissions of carbon dioxide, methane and nitrous oxide from Bed no. 5 at Helsinge STRB system. The x-axis represents the amount of days passed since initiation of a resting period. Emissions are given as mean values of four measuring stations. The bars represent standard errors. Notice different scales on the y-axes.

The flux chamber measurements also included N2O emission quantifications (Figure 3). Nitrous oxide is mainly produced by anaerobic denitrification (Firestone et al. 1980), but can also be produced by some oxygen requiring microbial processes (Lloyd et al. 1987; Robertson et al. 1995; Colliver & Stephenson 2000; Kampschreur et al. 2009). The surface emission of N2O increased as the resting period proceeded. Indeed, we found that N2O was mainly produced during the transition from anaerobic to aerobic conditions. This reflects that the sludge residue shortly after loading was dominated by anaerobic conditions (which were consistent with the results obtained from the probe measurements) and during the resting period gradually became more and more aerated.

DISCUSSION

Knowledge about the dynamics of gas compositions in sludge residue during a resting period is valuable, because it provides information on whether the sludge residue was dewatered sufficiently before more sludge is loaded into the bed. If the sludge residue during a resting period is not dewatered to a state where air is able to re-enter the pore space, permanently anaerobic conditions could develop. This would lead to an increased production of CH4 and N2O, which are potent greenhouse gases, having GWPs at 28 and 265, respectively (IPCC 2014). Aerobic mineralization processes are more efficient compared to anaerobic (Vymazal et al. 1998), in that sense that organic material is mineralized faster under aerobic conditions. Furthermore, insufficiently dewatered anaerobic sludge residue can cause odour nuisances (Nielsen 2005). To prevent such problems, STRBs are designed and operated to enhance aerobic conditions. The results of the probe measurements (Figure 2) showed that during the resting period, atmospheric air (represented by N2 and O2) re-entered the sludge residue and seemed to stabilize at certain levels at the different depth levels. During the resting period, the concentration of CH4 decreased at all depth levels, also reflecting that the sludge residue gradually became more aerobic. These results suggest that the bed used as the experimental site (Bed no. 5 at Helsinge STRB) was well operated. Methane was only produced under anaerobic conditions; however, CH4 was still present despite the presence of O2. When pore space air was collected from the gas probes, air was extracted from an area surrounding the slits in the probe. The presence of both CH4 and O2 suggests that the sludge residue was a mosaic of anaerobic and aerobic microenvironments.

The standard errors of the probe measurements (Figure 2) reflect the spatial variation between the four measuring stations. STRB systems are designed to enhance a homogeneous reed cover. Anaerobic conditions in sludge residue can, however, never be completely avoided. Reeds tend to grow in clusters if stressed by too anaerobic conditions in their growth substrate. The concentration of rhizomes in sludge residue underlying a reed cluster is dense, creating a vertical column down through the sludge residue, a so-called ‘dome’, dominated by aerobic conditions. This implies that the abundance of different gas species in the pore space of a vertical sludge residue profile can be rather different depending on whether the probes are mounted in a primarily anaerobic or aerobic area. In future studies, it could be relevant to investigate in more detail how the gas compositions of vertical profiles of sludge residue differ between anaerobic or aerobic areas and/or between areas with less and more dense reed clusters. It would also be relevant to relate the gas composition of different depth levels to physical and chemical characteristics (temperature, content of dry solids, NH4+ and NO3 etc.) of sludge residue from the corresponding depth levels.

CONCLUSIONS

The aim of the study was to measure the composition of gas species in the pore space at different depths in sludge residue in a bed of an STRB. The gas composition of the pore space in the sludge residue changed during the resting period. During the first part of the resting period, the CH4 concentration peaked at 10 cm depth. As the resting period proceeded, atmospheric air, and thereby O2, gradually re-entered the pore space from the surface. The sludge residue at 80 cm depth also gradually became oxygenated during the resting period due to the supply of atmospheric air from the reject water pipes. The sludge residue at 40 cm depth had a rather constant concentration of O2, suggesting that O2 at this depth was mainly supplied by rhizomes. The content of CO2 recorded at the different depth levels on the different days in the resting period seemed to follow the content of CH4, suggesting that the CO2 present within the layers of sludge residue was mainly a by-product from the CH4 producing process. However, the total emission of CO2 from the surface of the sludge residue exceeded the total emission of CH4, suggesting that CO2 was mainly produced in the top layer of newly applied sludge.

ACKNOWLEDGEMENTS

The presented work is a part of an industrial PhD project hosted by Orbicon A/S and The Technical University of Denmark (DTU). The PhD project is partly funded by the Innovation Fund Denmark, Orbicon and DTU. Furthermore, the authors wish to thank the staff at Helsinge WWTP for the possibility to use the STRB as an experimental site, and Nicole Brogaard Madsen (student at DTU) for undertaking the fieldwork.

REFERENCES

REFERENCES
Chanton
,
J.
&
Whiting
,
G.
1995
Trace gas exchange in freshwater and coastal marine environments: ebullition and transport by plants
. In:
Biogenic Trace Gases: Measuring Emissions from Soil and Water
(P. A. Matson & R. C. Harris, eds). Blackwell Science, Chichester, Sussex, UK
, pp.
98
125
.
Colliver
,
B.
&
Stephenson
,
T.
2000
Production of nitrogen oxide and dinitrogen oxide by autotrophic nitrifiers
.
Biotechnology Advances
18
(
3
),
219
232
.
Dominiak
,
D.
,
Christensen
,
M. L.
,
Keiding
,
K.
&
Nielsen
,
P. H.
2011
Sludge quality aspects of full-scale reed bed drainage
.
Water Research
45
(
19
),
6453
6460
.
Firestone
,
M. K.
,
Firestone
,
R. B.
&
Tiedje
,
J. M.
1980
Nitrous oxide from soil denitrification: factors controlling its biological production
.
Science
208
(
4445
),
749
751
.
IPCC
2014
Climate Change 2013: The Physical Science Basis: Working Group I Contribution to the Fifth Assessment Report of the Intergovernmental Panel on Climate Change
.
Cambridge University Press
.
Jeris
,
J. S.
&
McCarty
,
P. L.
1965
The biochemistry of methane fermentation using C14 tracers
.
Journal (Water Pollution Control Federation)
37
(
2
),
178
192
.
Kampschreur
,
M. J.
,
Temmink
,
H.
,
Kleerebezem
,
R.
,
Jetten
,
M. S. M.
&
Van Loosdrecht
,
M. C. M.
2009
Nitrous oxide emission during wastewater treatment
.
Water Research
43
,
4093
4103
.
Maljanen
,
M.
,
Kohonen
,
A. R.
,
Virkajärvi
,
P.
&
Martikainen
,
P.
2007
Fluxes and production of N2O, CO2 and CH4 in boreal agricultural soil during winter as affected by snow cover
.
Tellus B
59
(
5
),
853
859
.
Masciandaro
,
G.
,
Iannelli
,
R.
,
Chiarugi
,
M.
&
Peruzzi
,
E.
2015
Reed bed systems for sludge treatment: case studies in Italy
.
Water Science and Technology
72
(
7
),
1043
1050
.
Nielsen
,
S.
1990
Sludge dewatering and mineralisation in reed bed systems
. In:
Constructed Wetlands in Water Pollution Control, Proceedings International Conference on the Use of Constructed Wetlands in Water Pollution Control
,
Cambridge, UK
, pp.
245
255
.
Nielsen
,
S.
2003
Sludge drying reed beds
.
Water Science and Technology
48
(
5
),
101
109
.
Nielsen
,
S.
2005
Sludge reed bed facilities: operation and problems
.
Water Science and Technology
51
(
9
),
99
107
.
Nielsen
,
S.
&
Cooper
,
D. J.
2011
Dewatering sludge originating in water treatment works in reed bed systems
.
Water Science and Technology
64
(
2
),
361
366
.
Nielsen
,
S.
&
Willoughby
,
N.
2005
Sludge treatment and drying reed bed systems in Denmark
.
Water and Environment Journal
19
(
4
),
296
305
.
Obarska-Pempkowiak
,
H.
,
Tuszyska
,
A.
&
Sobociski
,
Z.
2003
Polish experience with sewage sludge dewatering in reed systems
.
Water Science and Technology
48
(
5
),
111
117
.
Olsson
,
L.
,
Larsen
,
J. D.
,
Ye
,
S.
&
Brix
,
H.
2014
Emissions of CO2 and CH4 from sludge treatment reed beds depend on system management and sludge loading
.
Journal of Environmental Management
141
,
51
60
.
Robertson
,
L. A.
,
Dalsgaard
,
T.
,
Revsbech
,
N.-P.
&
Kuenen
,
J. G.
1995
Confirmation of ‘aerobic denitrification’ in batch cultures, using gas chromatography and 15N mass spectrometry
.
FEMS Microbiology Ecology
18
(
2
),
113
119
.
Uggetti
,
E.
,
Llorens
,
E.
,
Pedescoll
,
A.
,
Ferrer
,
I.
,
Castellnou
,
R.
&
Garcia
,
J.
2009
Sludge dewatering and stabilization in drying reed beds: characterization of three full-scale systems in Catalonia, Spain
.
Bioresource Technology
100
(
17
),
3882
3890
.
Uggetti
,
E.
,
Ferrer
,
I.
,
Llorens
,
E.
&
Garcia
,
J.
2010
Sludge treatment wetlands: a review on the state of the art
.
Bioresource Technology
101
(
9
),
2905
2912
.
Uggetti
,
E.
,
Garcia
,
J.
,
Lind
,
S. E.
,
Martikainen
,
P. J.
&
Ferrer
,
I.
2012
Quantification of greenhouse gas emissions from sludge treatment wetlands
.
Water Research
46
(
6
),
1755
1762
.
Vincent
,
J.
,
Molle
,
P.
,
Wisniewski
,
C.
&
Lienard
,
A.
2011
Sludge drying reed beds for septage treatment: towards design and operation recommendations
.
Bioresource Technology
102
(
17
),
8327
8330
.
Vymazal
,
J.
,
Brix
,
H.
,
Cooper
,
P. F.
,
Haberl
,
R.
,
Perfler
,
R.
,
Laber
,
J.
1998
Removal mechanisms and types of constructed wetlands
. In:
Constructed Wetlands for Wastewater Treatment in Europe
(
Vymazal
,
J.
,
Brix
,
H.
,
Cooper
,
P. F.
,
Green
,
M. B.
&
Haberl
,
R.
, eds).
Backhuys Publishers
,
Leiden
,
The Netherlands
, pp.
17
66
.
Yu
,
R.
,
Kampschreur
,
M. J.
,
Van Loosdrecht
,
M. C. M.
&
Chandran
,
K.
2010
Mechanisms and specific directionality of autotrophic nitrous oxide and nitric oxide generation during transient anoxia
.
Environmental Science & Technology
44
,
1313
1319
.