The French connection
Measurement of greenhouse gas emissions from processes in five wastewater treatment plants, by researchers in France, has shown interesting results, says Frank Rogalla of Aqualia
Full-scale measurements have shown a large variation in N2O emissions between different wastewater treatment plants (WwTPs) and processes (0.001% to 55% of the nitrogen load emitted as N2O). First, denitrification was identified as the main N2O generation source, but more recent research has shown significant N2O generation during nitrification processes.
The main operational parameters identified as leading to N2O emissions are: n Low dissolved oxygen (DO) n High nitrite (NO2-) concentration during nitrification and denitrification and n Low chemical oxygen demand (COD)/N ratio during denitrification
Researchers from Suez Environnement's CIRSEE research centre took on-site measurements of N2O emissions, assessing N2O emissions from different biological WwTPs during 2007-8. A sampling and analysis method was developed and deployed to better understand and identify links between different processes and configurations, operation conditions and N2O emission.
N2O emissions from biological nitrogen treatment processes on five full-scale WWTPs were continuously sampled for periods of 4-6 hours and, depending on the process, longer periods (24 -72 hours).
The different processes consisted of one plug flow process, two conventional activated sludge (CAS) processes, one membrane bioreactor (MBR) process, and one biofilter process.
N2O air emissions were sampled on three types of emission:
- "Active" area source - presence of aeration flow: a square wooden air-sampling box was used to collect the airflow supplied by the aeration system, to quantify it (m3/h) by a gas mass counter and direct it to a gas analyser
- "Passive" area source - no aeration flow: an Odoflux dynamic flow chamber was used to sample the air at the surface of the tank, by covering it as hermetically as possible in order to isolate the surface from external conditions.
- Canalised source - outlet/pipeline: sampling was performed directly on the canalised air with N2O analysis equipment
Grab samples of wastewater were taken. The concentrations of different nitrous forms such as ammonium (N-NH4+), nitrite (N-NO2), nitrate (N-NO3-) were measured in the wastewater using standard methods. In addition, a VARIONPlus NH4/NO3 probe was used to continuously measure the NH4+ and NO3.
A mobile analysis system for N2O trace compounds was used to continuously analyse and record N2O in the sampled air. The Servomex infra-red gas analyser model 4210 has a range of 0-50ppm to an accuracy of 0.5ppm.
The measured concentrations were expressed in ppm volume/volume (v/v).
N2O emission surface flow was calculated based on the applied aeration flow (m3/h) and on the conversion of analysed N2O concentration in the sampled air (from ppm into mg/m3 air depending on gas flow and temperature. Related to the sampling surface area, the N2O surface flow could then be determined (mg N2O/h/m2).
Taking into account the surface area of the whole assessed treatment tank and the sampling time, the quantity of N2O emitted was determined and related to the quantity of Total Kjeldahl Nitrogen (TKN) that was removed by the overall treatment of the wastewater to establish a N2O emission factor (% of the removed TKN emitted as N-N2O).
Plug flow process
Two sampling campaigns were performed on the plug flow activated sludge process.
One consisted in continuous sampling of a fixed point in the middle of the aerobic zone for over 60 hours. During another campaign, 11 points along the whole aerobic zone were sampled enabling monitoring the same mass of wastewater following a "kinetic" approach.
- Overall, a high diurnal variation of N2O air emissions over the sampling period was observed, as found by previous researchers
- No N2O emissions were observed from the first anoxic zones where denitrification takes place
- Dynamic concentration changes in N-NH4+ and DO were observed, as is typical for a plug flow process, and which favour N2O production
- The N2O emissions were linked to the nitrification process, corresponding to recent findings - NH4+ concentrations decrease and N-NO3- and DO concentrations increase simultaneously. N2O is then emitted by stripping, linked with aeration flow
- A perfect correlation in time between the dissolved N2O production and the N2O air emissions was shown
- Related to the quantity of treated TKN, it was determined that less than 0.17% of the removed TKN was emitted to the air as N2O
On two conventional activated sludge aeration basins, continuous sampling was performed at a fixed point in the tank for about five hours and enabled measurement of the N2O air emissions over several altering aerated/non-aerated phases. During this campaign, dissolved N2O was not analysed.
As observed for the plug flow process, no N2O emissions were measured during denitrification (non-aerated phase). During the aerated phase, nitrification, relatively low N2O emissions were measured, following a two-peak emission trend:
- A first peak of N2O emission occurs at the beginning of the aeration phase when the blower gets started. This is probably due to the stripping of the N2O previously generated during denitrification in the anoxic zone
- A second peak of N2O emission occurs during the course of aeration, indicating that the N2O emissions are linked to the nitrification process (stripping of N2O produced during nitrification).
Similar results with the same two-peak N2O emission pattern was detected at each aeration phase, when ammonium was present, on both plants, and overall the measured N2O emissions from the CAS processes were low. Related to the quantity of treated TKN, it was determined that less than 0.01% of the removed TKN is emitted to the air as N2O, more than 10 times less than compared to the plug flow process.
While important concentration changes would cause high N2O emissions as discussed above, the well mixed hydraulic conditions with high recycling rates typical for a CAS process have been identified as leading to a much lower N2O generation.
In a membrane bioreactor (MBR) process, N2O emissions were assessed in the aeration and ultrafiltration (UF) tanks. The aeration tank was continuously sampled for 48 hours and results for a period of nine hours.
N2O is first measured in the morning, corresponding to the arrival of ammonium to the WwTP. These results clearly show the diurnal variations of the N2O emissions linked to the incoming nitrogen load, as shown by other full-scale studies. These results also show that the N2O emissions correspond to the aerated periods in the tank and follow a twopeak emission pattern as observed for the CAS processes.
The UF tank was continuously sampled for 48 hours and N2O emission results for a period of nine hours were recorded. Taking into account the quantified N2O emissions from the whole MBR process and the quantity of TKN removed, it was determined that 0.11% of the removed TKN was emitted as N2O, with about 60% of the N2O emissions measured on the aeration tank.
Measured concentrations of N-NH4+ and N-NO3- and DO in the aeration tank indicate the occurrence of incomplete denitrification that could partly explain these N2O emissions.
The assessed WwTP with biological active filters (BAF) is totally covered. Hence, N2O emissions from the odour treatment unit represent the global N2O emissions from the plant and were continuously sampled for 72 hours.
The N2O emission peaks are clearly linked to the washing periods of the filters, when the high airflow into the biofilter expands the granulate support media and leads to a stripping of N2O produced in between backwashes (see figure 5). N2O emissions from the first phase biofilters designed for carbon treatment and from the second nitrification phase were continuously sampled for 24 hours.
The measurements on the top of the biofilters showed higher N2O from the first phase, indicating partial nitrification during limited DO concentrations caused by competition for oxygen by organic matter. The lower N2O emissions from the nitrifying biofilter could be explained by a lower nitrogen load and higher DO concentrations. Taking into account the quantified N2O emissions from the whole BAF process and the quantity of TKN removed, it was determined that 0.6% of the removed TKN was emitted as N2O.
Notorious nitrous oxide
The full-scale N2O air emission measurements carried out on different biological nitrogen treatment processes on five WwTP's showed N2O emission factors below 1% of the removed nitrogen. Besides the variation of N2O emission factors between the different plants, the longterm continuous sampling showed temporal variations of N2O emission within the biological tanks.
The emission profiles presented for the different assessed processes showed that some N2O can be produced during non-aerated (anoxic) steps, but it is always emitted during aerated steps (stripping). Moreover, it was shown that N2O was mainly produced during nitrification. The results suggest that the key parameters of influence on N2O generation are linked together and comprise high ammonium concentration and limited DO concentrations in the biological tank, most probably linked to nitrite accumulation.
From current modelling of the overall GHG balance of a plant, the CO2 equivalent emissions from carbon oxidation, including respiration and BOD removal, are about 0.4kg CO2e/m3.
An equivalent amount compared with the biological treatment is estimated to be generated from sludge handling, adding digestion and biogas conversion to electricity, around 0.25kg CO2e/m3, to sludge disposal and reuse, at 0.2kg CO2e/m3.
Compared with these two main sources that add 0.85kg CO2e/m3, the N2O emissions at 0.2kg CO2e/m3 are equivalent to another 25% , although the latter can vary significantly depending on the nitrification process and its control.