 Hello everyone look it started now I'm Jose Poro cobalt water global I'm pleased to welcome you to our masterclass webinar series in conjunction with the launch of the book Quantification and modeling of fugitive greenhouse gas emissions from urban water systems Which is published by IWA publishing and was lead edited by Liu Ye of University of Queensland and co-edited by Ingmar and Opens of AM team and Ghent University and myself Today's masterclass, which is the second of four is on monitoring modeling and mitigating wastewater nitrous oxide Process emissions. We have an excellent lineup comprised of some of the book contributors but before we jump into things I want to thank IWA and I can me for hosting us and Amanda Lake of Jacobs for taking the initiative to help spread the word on the book and through this masterclass series but more importantly for helping to inspire action and By engaging the water community with the content of the book Thank You Amanda for your leadership and passion for climate impact so getting back to N2O when it comes to N2O what is interesting is that We're essentially living in two realities And I was explaining this to an innovative water utility a couple of weeks ago We have one reality, which is I consider to be an alternate reality because not a consistent with the science or based on truth and the other Which is the one that is based on science and in seeking the truth about N2O because In the alternate reality we're using generic emission factors like those like the IPCC emission factors Which are not capable accurately accounting for N2O because of its extreme Because it is extremely site-specific But recognizing that this was one of the key motivators of the book wanting to provide tools and methods for better Getting to the truth about N2O and if we can get to the truth on N2O we can reduce N2O So let's get things kicked off. It's my pleasure to introduce to you your moderator Dr. Maite Pijuan who is the head of technologies and evaluation research area at the Catalan Institute for Water Research or ICRA It's worth noting that Dr. Pijuan is quite the N2O expert herself and has contributed a couple of the few of the chapters that helped contribute a few of the chapters in the book. I've had the pleasure and honor of working with her on several collaborations and What's interesting is to point out is it was exciting because she came up with a way to an Alternate an alternative method for determining the emissions from non aerated zones and this is in the book So without further ado Maite Floor is yours. Thank you Thank you Jose for the introduction and welcome everyone for me. It's a pleasure to share this webinar on N2O monitoring modeling and mitigation in N2O I'm sure we will learn a lot We have four fantastic panelists three of them from the academia that have contributed to this book and another one a special guest from an industry from Australia so Next slide please Before going into the presentations just to let you know that this masterclass series is being organized by the IWA Climate Smart Cities Group Which is a platform aiming at knowledge exchange with two active subgroups So if you are part of the IWA EWAC means the members and you're interested in this topic Please consider joining these activities of this group through the IWA connect site please next There's some practical Information of the webinar that will be recorded and made available to everyone free of charge through the IWA website Next please Also some practical information We will have first four presentations with a brief Q&A section after each of them And at the end we will have also discussion with a Q&A Section all the attendees you have the microphones Muted and we cannot respond to raise hands. So if you have questions, please raise them by using the Q&A box We'll try to answer them along the way in the chat or later on in the Q&A section Also, if you have a general request that are not specific to the topic of the webinar, you can use the chat box please next That's a little bit of the agenda. We'll have These four panelists that he was mentioning with 10 to 15 minutes presentations and then a short Q&A discussion At the end of each talk and then one or two days longer at the end of all the four speakers And then we will end up with the final remarks and conclusions next slide, please Well, just a picture of us next please Also we'll have a poll We want to know where you guys work and Then that utility you work for have a net zero target and also what action are you taking on On try to reduce this process emissions. I believe this poll will be open for all the the webinar until the end Next slide, please Also, if you share your thoughts on this topic on social media, please stack us and don't forget to include the hashtags we are mentioning and We are moving now to the to the first Panelist it's a pleasure for me to introduce Dr. Vanessa Paravitsini She's a researcher at the Institute of Water Quality and Resource Management and TUV in Vienna in Austria And her research topics include industrial and municipal with water treatment in house gas emissions from with water treatment plants Life cycle assessment in urban water management and biofilm and with water treatment systems And she is the leading author of chapter five of the book that focuses on full-scale Quantification and GIG emissions from your one with water systems and in her presentation She will explain as a different available Methodologies to reliable quantify and to a mission from with water treatment plants, please Vanessa the forlorn is yours Thank you Marta for the introduction. Hello, everybody Well, I just check if I can control the slides. Yes so Well when aiming to mitigate and to emissions at which water treatment plants it is imperative to understand the mechanism that Trigger its formation Therefore my first slides will focus on the pathways of and to a generation and in the second part of the presentation I will focus more on the quantification methods for and to emissions at with water treatment plants The majority of and to a production originates from microbial mediated nitrification and denitrification processes occurring in both terrestrial and aquatic systems and being also We know key processes in the biological treatment of wastewater So during nitrification here on the left side Ammonium oxidizing bacteria which perform the oxidation of ammonia to nitrite are known to be able to produce large Quantities of and to as a byproduct of their metabolism So two pathways have been postulated based on extensive studies. The first one is the hydroxylamine oxidation and the second one is the autotrophic nitrite reduction also called Nitrify at the nitrification, but I think you will hear more about this to pathways in in the next presentation I just would like to say that the first pathway seems to be favored under a high ammonia oxidized oxidation rates and the second one might be predominant under limited dissolved oxygen concentrations On the other hand on the right side It's a trophic denitrification has the potential to produce and to consume and to all and to all is an obligatory Intermediate in this process. It's not a byproduct and being produced along the multiple reductions that Starting from nitrate and leading to nitrogen gas So when the process performs completely and to all will not accumulate and the denitrification will act actually as a sink of and to all Having the potential of reducing also the dissolved and to all previously produced by other processes like has the nitrification However, in some cases this last step from Of the denitrification of and to all to nitrogen gas And my may not be as efficient as the previous one and Because it's known that this step is a more sensitive to environmental factors So in these cases and to all will accumulate and then can Release to the atmosphere. So the take home message here is for me Nitrification is a source of and to all and denitrification can be a source but also a sink for into Based on these fundamentals, it can be expected that the main sources of and to all emissions in mainstream with water treatment Processes will be the units performing the nitrification and nitrification with and to all being produced from both anoxic and aerobic zones But the aerobic zone, of course, have been reported to contribute more to the emission of and to all being this promoted there by the Irrational stripping In comparison In novel nitrogen removal processes over partial nitritation and anamox process if more intensive and to all emissions were found from the aerobic zones In the following anamox process, however, the anaerobic ammonium oxidation was not reported to produce and to all Several operational and environmental conditions are found to impact and to a generation and emission during nitrification and denitrification including loading conditions Dissolved oxidant concentration, pH value, temperature and so on And the same applies also to the substrate and intermediates of the nitrogen removal processes Such as the concentration of ammonia, of nitrite and during the denitrification the availability of organic carbon So the prediction on how Entry or generation is influenced influenced by the different parameter is a quite challenging task and As different conditions are applied simultaneously during the wastewater treatment process But all in all it can be said that process conditions supporting extensive nitrification and denitrification usually lead to lower and to all generation and emissions rate The high number of triggering factors Impacting and to a generation in wastewater treatment plants is the main reason for the significant temporal and seasonal variability Detected in measurement surveys up to date and to describe these dynamics A continuous online monitoring is a must and this is the reason why the full-scale quantification of N2 Emissions required significant input of resources on site. So we are already moving To the topic full-scale quantification of N2 in wastewater treatment plants and of course, this is quite important because a key to formulating strategies to control and reduce N2O emissions A key is the identification and the quantification of all the sources we have at the plant And so in the next slides, I would like to give a very short overview on some methods that can be applied for the quantification of N2O emissions at which was a treatment plant You will find in the book a much more detailed descriptions of the methods Well, first of all, we have several methods. So how to select the source of the N2O emissions and how to select the source of the N2O emissions We have several methods. So how to select the most suitable one. The selection of the suitable quantification method is mainly dictated by the objective of the measurement Being the compliance with greenhouse gas emissions protocols or the calibration validation of prediction models or the development of mitigation strategies So in each case, a different degree of information is required and these need to be considered in the selection of the method Quantifying methods can be classified into plant-wide and process unit measurement approaches. Plant-wide quantification enables the determination of the overall N2O emission as the plant, including sources that might be missed when using a process unit method or might not be accessible In contrast, the process unit approach identifies and quantifies single N2O emission sources, allowing a much deeper understanding of the mechanism of the N2O production and emission and how they are linked to the process parameter Yes, one method for the plant-wide quantification that has been already applied with water treatment plants is the mobile trace against dispersion method This ground-based remote sensing method uses a controlled release of a tracer gas positioned at the plant, combined with measurements of atmospheric gas concentrations taking downwind of the target area, in this case the wastewater treatment plant So when the tracer gas is released at the constant rate from the emitting area, N2O emission rate can be then calculated in real time by relating the measured plume transverse concentration of N2O and of the tracer gas So for the measurement, you need a mobile measurement platform, usually a vehicle equipped with the analyzers and a global navigation satellite system for recording the measurement locations Among the process unit quantification methods, the floating hoot method is probably the most common approach to sample the off-gas, leaving the surface of activated sludge tanks with bubble aeration This can be done with one or multiple hoots, according to the local requirements on site And the off-gas stream captured by the hoot is usually fed to online gas analyzers for the quantification of N2O And to calculate the N2O flux out of the activated sludge tank, you also need the off-gas flow rate of the tank So this is also quite important input for the estimation In the liquid to gas mass transfer method, the approach is different Here the N2O concentration in the liquid phase is measured using for example a dissolved N2O probe And the two N2O emissions to the gas phase is then estimated using a gas liquid transfer coefficient So the KLA coefficients of N2O across the gas and liquid interface This KLA value can be estimated, applied at theoretical or more empirical approaches And this calculation is not that trivial What is important to underline is that in both methods, it is quite challenging to extrapolate the punctual measurement result to the total emission of the activated sludge tank Well, in chapter 5, we put much effort in providing a general overview on the methodologies currently available for the quantification of N2O emissions At least with the treatment plant, of course And well, highlighting also their field of applicability, the instrumental requirements, the strengths and limitation of each method And recommendations in regard to supplemental data requirements and the duration of the measurement campaign are also given So in conclusion, it can be said that significant improvement has been achieved in the quantification of N2O emissions at waste water treatment plants in the past decade And proceeding from grab sampling to online monitoring from short-term to long-term measurements campaigns from process units to plant-wide quantification methodologies However, quantifying N2O emissions still remains a challenging task due to the special and temporal variability of the emission pattern, which are strongly influenced by the environmental and process conditions So in this context, a widely implemented measurement protocol would help improving the quality of the data generated in these campaigns and the comparability of the result will also be easier At the moment, this widely implemented measurement protocol is still lacking And additionally, I think further efforts are necessary for linking N2O emissions with plant performance indicators and operational conditions because this link is the key for developing effective mitigation strategies So in the end, I would like to acknowledge the authors of the book chapter 2 and 5, book chapter 2 focused on the generation mechanism of N2O and the chapter 5 focused more on the quantification methodologies So thank you very much for your attention Thank you very much, Vanessa, for your talk. I think it was a very nice overview of the mechanisms of N2O and also the different ways we have to quantify these gas emissions from waste water treatment systems I believe we have one question from the chat that was referring more on how you can quantify the flux coming from the anoxic zones If you want to comment on that, like with the HOOT method, how do you get the flux that comes from the anoxic tanks? Well, let's see. Of course, you are monitoring the concentration in the HOOT, in the headspace of the HOOT, so you can see according to the changing of the concentration there if you have a production and also emission in the headspace of the HOOT So you could use, let's say, this somehow static measurement method to get there and let's say an estimation I mean, we know that what is emitted during the nitrification phases is not very as it's compared to what is emitted during the aeration phases is usually not that relevant in the whole balance Of course, it's also interesting to understand what's happening during these phases and if you parallel to the measurement with the HOOT, you also use in the liquid phase a probe, an actual probe, then this can also give you some more information on what's happening in the water Because sometimes, of course, you cannot see immediately in the HOOT what's happening. If you have an internal loop in the HOOT and I mean, you can also use some sweep gas also maybe to have some circulation in the HOOT that can help Yes, of course Thanks Vanessa. Just to compliment on that, I have seen two methods one was using the tracer test, tracer gas, which is what you mentioned. In our experience, we had an analyzer that was able to measure also oxygen, so we were using that as a sweeping gas through the HOOT And then we knew the concentration of oxygen in the air and then the concentration of oxygen coming out of the HOOT and we could estimate the flux coming from these anoxic zones since they don't have any oxygen, so in a way it was similar to a tracer gas And then it was also a common, maybe the last one for the short Q&A session, that was how to make sure the section you covered with the HOOT is representative from the This is a very good question The place you are measuring since the tanks are very big and the HOOTs are not so big Yeah, that's a very good question and yeah, of course, you have really to think about it I mean, if you have, let's say, a tank which is quite a good mixed, I don't know, a tank with a circulating flow or really as a completely steered tank reactor Then probably you will not see a very high concentration gradient in terms of ammonium or nitrate So, I mean, of course, we usually do, it's a must at the beginning of a campaign to have some to monitor at the different points and see, compare if you have a, let's say, a comparable concentration or emissions And so from our experience, if you have a quite well mixed tank, then you will not see very much differences But if you have a big gradient in the concentration, so a plug flow reactor, so if you're also measuring several cascade Then you are really to think about where you put the HOOTs and sometimes it's a good idea not to use just one HOOT Yeah, fully every... I'm not sure if we lost Vanessa, but anyway, we have to move to the next panelist I couldn't hear you anymore We lost you just for a second, but no problem We can further discuss that later on at the end of the talks But now we need to move to the next panelist, that is Dr. Haran Dwan He's a research fellow at the Australian Centre of Water and Environmental Biotechnology For many of us, many know as the former Advanced Water Management Centre at the University of Queensland And his research focus on novel wastewater treatment technologies, nitrous oxide emissions and sludge management And he's the co-author of several chapters in this book, so thanks, Haran You have been very active in helping to get this book out And today he will tell us all about this fantastic wall of anti-wall modelling and mitigation strategies, the platform is yours Thank you, thank you, Mighty Hello everyone My presentation will be further to Vanessa's presentation on monitoring and quantification of anti-wall emissions from wastewater treatment plants As Vanessa has introduced the production mechanism of anti-wall is quite complicated Governed by many different environmental factors So if you vary one, you may not see direct consequence on the anti-wall emissions But may have other relevant environmental factors changes So really to understand anti-wall emission mechanisms we will have to extract the mechanisms into mathematical modelling So we can better understand the regulation of anti-wall emissions in wastewater treatment plants With such modelling tools available, we will be able to more confidently develop mitigation strategies to optimise our process design or process operation for less anti-wall emissions Yes, so today I will be presenting mainly our Chapter 7 on modelling and tool production emissions This chapter is led by Professor Matthew Spenandio from INSAR France And in addition, some modelling relevant mitigation strategy study from a work that I did with some collaborators So if you are not that familiar with anti-wall modelling, anti-wall model is an extension of activated sludge model So it's developed from activated sludge model ASM And ASM is essentially governed by two things Reaction kinetics, which is also called monokinetics On the right hand side you will see two equations Those two equations governs the consumption of substrates And as well as as a result of substrate consumption, the growth rate of microorganisms Those two equations are highly correlated And essentially if we knew all the consumption of substrates in the wastewater system We expand all these equations and we can model all these reactions with reaction stored geometry Down there we see a matrix is a good example and also a simplified example of modelling If we look at the first row, which we see the process rate Which is just exactly the same as what we see up there in equation 1 and 2, governs the process rate And with reaction stored geometry, we will see how much substrate is consumed For example, SO represents oxygen consumption And how much microorganism is grown as a result of this substrate consumption And essentially mathematical modelling is a mathematical representation of wastewater treatment processes that we understood So now with advanced knowledge on until production mechanisms We will be able to extract least understanding to build mathematical models So now I will quickly walk you through the development of many end-tool models Just to refresh your memory a little bit, this is just a simplified version of until generation illustration A simpler one that many have just presented So we have mainly three main biological production pathways of end-tool Two by ammonia oxidizing bacteria, AOB, namely nitrified de-nitrification pathway and hydroxylamine oxidation pathway And we also have heterotrophic de-nitrification pathway carried out by heterotrophic de-nitrifying bacteria As for AOB, initially there were four different single pathway model proposed Just based on the mechanism that we understood Model A and B on the right hand side shows two nitrified de-nitrification pathway model With only minor difference whether hydroxylamine was included in the conversion or not And similarly, for hydroxylamine oxidation pathway modelling There were two models proposed and also with small differences on whether an O is included or an OH is included And intuitively, since those single pathway model were developed And researchers later on combined them together And the model on the upper side shows a consolidated two pathway model still following the model kinetics And an alternative model was also proposed since nitrified de-nitrification pathway takes electrons So it was proposed there may be some electron carrier shortage or it's regulated by electrons So in this model proposed by me at UQ, they proposed electron carriers in the model to regulate the two pathway And these two pathway model were later on validated by isotope studies And showing the prediction of pathway contributions are very similar to what was measured by isotope studies So these A will be single pathway and two pathway models have its applicabilities So single pathway model obviously is only applicable when one single pathway is dominant under some environmental conditions Whereas in the environmental conditions where none of these pathway dominant is that both of them exist Then only two pathway model would be accurate to describe the two emissions under such scenario So similarly after the A will be pathways for anti-emission from de-nitrification pathways There were three models developed And first one being the classic ASMN model Basically in this model de-nitrification each steps of de-nitrification were modeled from nitrate to nitride to an O to an N2O then an O So we will be able to get an N2 emission from these comprehensive stepwise modeling Then later on it was realized there were actually electron competition in de-nitrification process Basically some sub steps may be more preferred for electrons So Pan et al at UQ proposed electron carriers in de-nitrification model called ASM-ICE Basically in this model electron carriers were added in More recently an alternative model was proposed by Batzmeier's group at DTU They used an electric circuit to resemble the electron competition Basically using electric resistance to represent each steps to regulate the electrons that became available And how that is regulated the N2 emissions in de-nitrification And with all the single pathway or two pathway models proposed After all it can be then unified to build a consolidated model Which includes all major pathways so that it can be used to describe N2 emissions at real wastewater systems So this is just an illustration of our previous work that we did for N2 mitigation And we can see mathematical modeling plays a critical role in N2 mitigation And then our next speaker will elaborate more on this So I will just skip this And with all our understanding in modeling Now we start to look at how N2 mitigation is governed There has been many N2 mitigation studies in lab scale, pilot scale and some recently also at full scale There are just too many to be summarized And then we dig into the depths and we start to from a retrospective point of view And we see it's all actually from the modeling With the modeling that we understood that the generation pathways open to all And we understand how these mitigation strategies were developed So if we again look from N2 generation pathways If we look at the hydrocellular oxidation pathway Effectively we knew that N2 emissions from hydrocellular oxidation pathway is related to the AOR Ammonia oxidation rate which can be described in mathematical model in this equation So we can see that if we are able to reduce AOR without compromising nutrient removal performance We will be able to reduce N2 emissions So here we can see from this equation If we are able to reduce the ammonia concentration in the reactor Or if we are able to reduce oxygen concentrations We will be able to reduce N2 emissions from this pathway And this is exactly why some studies did low DO or flow equilibrium To reduce ammonia concentrations And similarly for N2 generated from nitrified denitrification pathway It is governed in modeling by this equation So similarly we will be able to reduce nitrified concentrations to reduce N2 emissions from this pathway And also since this pathway uses electrons generated from hydrocellular oxidation So essentially if we are able to reduce the electron supply from the other pathway There is also a potential to reduce N2 emissions from this pathway as well And this similarly applies for denitrification pathway as well With a difference that in denitrification there is a pathway for N2 generation Basically from an O reduction as well as a sink for N2 Namely the N2 reduction pathway So if we are able to regulate the environmental conditions or the operational parameters To regulate an O reduction And maximizing N2 reduction that we will be able to reduce N2 English water treatment systems And as for the physical process of liquid mass transfer If we are able to reduce liquid to air transfer we will also be able to limit N2 emissions from liquid phase And finally end of pyro treatment technology such as capturing the gas Using bio filters to reduce N2 emissions As I said there has been a lot of mitigation strategies and proposed and developed And in recent years many have now been tested and demonstrated at full scale Showing some examples here Like the first one was actually carried out by MIT school They changed continuous aeration to intermittent aeration at a full scale SPR plan And resulted in 90% reductions in N2 emissions And the mechanism is from multiple pathways And similarly there has been other attempts in total sink for and more expected to come Since mitigation or the modeling is really reaching a maturity And so finally take home message N2 generation mechanism has greatly progressed in recent decades So that the N2 models were developed with these mechanisms In general these are very similar But with some different mathematical structures N2 modeling has reached maturity that has been applied at full scale to estimate N2 emissions And guide N2 mitigation strategies development Integration of N2 models into a plant-wide model could be a powerful tool for future optimization works And finally N2 mitigation is also reaching a maturity And some have been demonstrated at full scale But still more demonstrations are required And I want to acknowledge Professor Trico Yuan and Social Professor Liu Ye at UQ Who are leading the Queenhouse Gas Research at the University of Greenstone Also want to acknowledge Professor Matthew Spirna Ndiel Who is the lead author of chapter 7, the modeling chapter I also want to acknowledge our collaborator George Wells and Conrad Koch Who were co-authors of the mitigation paper Thank you Thanks Haoran for this very indeed presentation on models And many questions there I haven't seen any specific question on your talk in the chat But I will just make maybe a general one to you And maybe we can further discuss it at the end So my understanding is it's very, I'm not sure at this stage But it's very complicated to correctly model these emissions They are variable across the day, they are variable across the year One pathway, multiple pathways, every plant may behave in a different way Plus we have the adaptation of the biomass that is adapting to different stressors That they can have different emissions depending on the plant So from your experience, what kind of experimental data do we need from the plant? We want to model to ensure that we have good model predictions So should we start plugging into sensors in the plant in the SCADA to get this data as well And to have how much data do we need? Do we need one week of data? Do we need a week per month during one year? What is your feeling? Yeah, that's a very good question Yeah, untold emissions is highly dynamic So in terms of the monitoring duration, obviously one week or a few days wouldn't do the work Many people who did long-term monitoring have repeatedly observed seasonal variations of untold emissions So if you want to monitor untold, if budget allows, it's better to do over a year monitoring to capture whether there is seasonal variation of untold emissions So that's a question for duration And then for the variability of untold emissions in treatment plants and how to model them Yes, modeling could be a big challenge at FOSCADEL if you have changes So if we're talking about some theoretical ground, if you don't vary your operation and your plant is operating smoothly And similar to what you experienced, then the model should still hold and the pathway theoretically shouldn't change But with for treatment plants it's complicated, everything could happen And that is something we still need to look at And then consequently online monitoring or monitoring untold as part of with for treatment plants routine monitoring would be helpful to address such uncertainties Thanks, I can see that there is one question in the chat now When it comes to mathematical modeling, is there a way to get around the computational limitations when it comes to full-scale modeling? Is there a way to get around what? The computational limitations Computational limitations Maybe if the person that did make this question can further explain what Xuxi is referring to and we can maybe discuss at the end Yeah Because now, yeah, I think we need to move to the next one Thanks a lot, Auran, and we'll see you at the end for the final discussion Yeah Our next panelist is Dr. Ben Banden Acker I hope my pronunciation is okay Ben is the Lead with Water Scientist at the South Australian Water Corporation He's also joined senior lecturer at Flinders University in South Australia and a joint associate professor at the University of South Australia He works on different research projects with the goal of improving the environmental performance and sustainability of wastewater treatment And he also seeks to improve public health outcomes of water reuse schemes across South Australia And in his presentation he will tell us the Antwoa monitoring campaign conducted in the largest wastewater treatment plant they are operating in South Australia And how that helps to implement mitigation strategies Yeah, then the floor is yours, thank you Thank you very much Okay, here we go. So just to provide some background on from South Australia we are owned by the government of South Australia We are responsible for providing water and sewerage services to more than 1.7 million South Australians And scope 1 emissions from our 24 wastewater treatment plants represent around 22% of our total corporate emissions And these scope 1 emissions are set to increase to 65% by 2030 due to greening of our electricity grid So in the last 15 years South Australia's energy supply has transitioned from 1% renewable to 60% And in the coming years renewable energy will reach around 85% So a lot of our focus will be on our scope 1 emissions such as nitrous oxide These emissions are currently estimated and we report these under a national greenhouse and energy reporting scheme We have a target of net zero by 2030 and to achieve this target it will require direct monitoring to understand what our true emissions are And direct monitoring will also be required if we want to explore the role that process optimization will play in reducing our scope 1 emissions Now the journey of monitoring our scope 1 emissions started in 2012 when during the introduction of a carbon tax here in Australia which was later revoked The focus for this work was on Bolliver wastewater treatment plant. This is our largest wastewater treatment plant It actually has two activated sludge reactors. It has the step fed activated sludge reactor you can see at the top And it also has a smaller sequencing batch reactor. When it comes to our activated sludge reactor, our step fed reactor What happens is we get spatial gradients in many parameters, you get spatial gradients and dissolved oxygen and nitrogen concentration and biomass concentration And these can impact nitrous oxide production in different ways So the focus of this work was trying to capture this spatial variability in nitrous oxide emission as well as the diurnal variability And this is of course required in order to enable accurate quantification of our emissions But also in doing this you can identify hotspots along the reactor and then identify potential causes for those higher emissions And this information can then be used to target control measures. How can we make operational or design changes to the plant to try and reduce these emissions So how did we quantify variability? Well what we did was developed a multiple gas hood system So we developed a multiple gas hood system and that enabled I guess spatial and diurnal monitoring So we were able to use multiple floating gas hoods, so we had a number of these floating gas hoods That would pipe the captured off gas to a centralized greenhouse gas monitoring unit We then from each hood we were able to measure temperature and pressure and flow rate And we used the PLC controller to open and close gas solenoid valves to direct a portion of the gas from each hood to a single Hariba infrared gas analyzer So every five to six minutes the analyzer was able to measure nitrous oxide emissions from multiple gas hoods that were located along the activators sludge reactor And this is what it looked like in practice so at the top we were able to monitor from six locations from six hoods positioned along the activators sludge reactor That's our stepbed reactor and at the bottom we are able to monitor nitrous oxide from our sequencing batch reactor from three separate locations And if you're after more information about the methodology that we applied here you can see that in the paper that's been published here What we also did was undertake intensive monitoring of a whole range of different water quality parameters that are listed here And the aim of this is to try and identify relationships between these parameters and nitrous oxide to try and understand the production pathways that are responsible for nitrous oxide production But also this information was also important to support and help calibrate the development of a multi-path way model which Haran was previously talking about In terms of key findings the graph on the right shows a diurnal profile of nitrous oxide measured from the six different locations in our activators sludge reactor What you can see is there is pronounced diurnal variability but also large spatial variability along the reactor And this highlights the importance of the need to try and monitor for multiple locations to try and incorporate and understand and characterize this variability Once we incorporate this variability what we found was that the emission factor for this particular plant was quite high Almost 2% of the nitrogen going into the system was emitted as nitrous oxide and that equated to around about 30 kilotons of carbon dioxide equivalent per year And when we compare this to the use of a generic emission factor to estimate our emissions we actually reported 9 kilotons So you can see there's a large discrepancy between what we measured and what we actually estimated using a generic emission factor But in doing this we're able to identify the causes and we're now taking steps to look at how we can mitigate this And I'll talk about that in the next few slides In terms of our sequencing batch reactor despite being a well mixed system we still identified that there was spatial variability in nitrous oxide And again it stresses the need to make sure you capture this spatial variability from monitoring at multiple locations The emission factor that we measured was close to 0.9% which is what we usually expected and typical of this type of system But we did identify that the way that the plant was aerated aided in nitrous oxide production and we took steps to try and mitigate that So looking at ways to mitigate we looked at ways to mitigate and try and reduce our nitrous oxide emissions with the support from a multi-pathway nitrous oxide model Which Hurran previously talked about so with the support of the University of Queensland We were able to calibrate this model using long-term nitrous oxide data and also long-term process data as well And this model was really good at predicting emissions and also providing insights into the types of pathways that are probably responsible for the emissions that we observed Using this model it's great to be able to then customise design and operational changes So what sort of changes can you make to the plant from an operation perspective or what sort of design changes can you make And how do you make these changes and allows us to assess these changes without actually compromising the performance of the plant and also Increasing the cost of treatment as well So just to give one example the first is looking at operational improvements to our sequencing batch reactor We noted that the aeration profile so at the bottom in the brown there we've got our dissolved oxygen profile This aeration profile in the brown actually aided nitrous oxide production via the hydroxylamine oxidation pathway And also from a nitrified denitrification pathway The model was then able to look at different DO profiles So we looked at the profile strategy, the DO profile in the red and also the blue And we identified that by changing the DO profiles the model identified that we could achieve around about a 35% reduction in nitrous oxide emissions If we modified the aeration profile What we did is we went for strategy two and with fine tuning of our dissolved oxygen set points at this plant We implemented these changes and we went out and revalidated and measured nitrous oxide And we found that around about 30 to 35% reduction in nitrous oxide just these simple changes to the dissolved oxygen profiles The reductions that we measured in nitrous oxide emissions were also consistent with the predictions of the model as well The other important point to make here is that in doing this we also identified additional operational and performance benefits Not only did we observe a reduction in emissions but we also were able to observe an improvement in total nitrogen removal We went over aerating the system so we were able to achieve about a 20% improvement in total nitrogen removal And we also noted an improvement about 20% improvement in reduction in energy that was associated with aeration as well So not only did we achieve a higher level of nitrous oxide reduction but there was also additional operational performance improvements And if you're after more information please refer to this paper that we published which covers this topic in much greater detail We're also looking at design improvements considering design improvements to our step-fed activated sludge reactor What we found was the high emissions that were particularly noted in the second step-feed of the reactor was actually caused by an uneven biomass loading rate within this reactor And the reason for this was it was caused by the location of the return activated sludge So the return activated sludge is returned to only the start of the reactor in the first step-fed section And as a result the mixed liquor concentration as it goes along the reactor becomes diluted by the second step-feed So the mixed liquor concentration in the second stage of the reactor is actually much lower and as a result the biomass loading rate in the section was much higher And through the support of the model we're able to undergo, you know, assess different scenarios, we assessed what would happen if we were able to return some of this RAS to the second step in order to try and even out the biomass loading rate across the reactor And what we found was looking at different scenarios if we returned about 30% of the RAS to the second step It showed potential to cut our nitrous oxide emissions by around about 16 kilotons of carbon dioxide per year, so quite significant potential with these sort of design changes And again if you'd like to see more information and know more about this work and this model, I'll refer you to this publication that's presented here So just to finish, I guess the take home message here is I guess the importance of good measurements and accurate data It really does require the need to capture variability, so in order to have good measurements and good data it's important to capture variability We've showed examples where there is pronounced variability in emissions along our reactors And in doing this we found that there is large discrepancy between the use of a generic emission factor versus what we actually went out and measured And when you set out to capture this variability you're able to identify hotspots along the reactor and this can then help us target better target control strategies The other benefit of having good measurements is you can generate an accurate model and use this model to try to understand and validate the causes for nitrous oxide emissions And then using this model you can customize and test different control strategies in a way that you don't compromise the functional performance or the cost of wastewater treatment And finally, another final conclusion is that from our work operational improvements can play a major role in reducing emissions from the work that we did on our sequencing batch reactor And in doing so additional benefits can also be gained in terms of cost reduction and also other benefits to improvements in the operational performance Before I finish I'd also like to acknowledge the contribution of a number of members from the University of Queensland and also Peter Barr from SA Water From our Water Engineering Technologies who developed the online multiple gas monitoring system that we employed at Bollival Wastewater Treatment Plant Thank you very much Thank you very much and this is a clear example of a nice collaboration between academia and industry and the results are just fantastic I mean sometimes we get some reluctancy in trying to establish the real emissions of our wastewater treatment plant But here you show how that can also have benefits like economical benefits you obtain like a reduction on the aviation and a reduction of costs and a reduction of emissions so that's so fantastic We have many questions for you in the chat so we'll maybe can answer here a couple of them One is how does reducing the ideal concentration impact the electrification process as I'm assuming well here it says that the work needs to meet the final effluent ammonia concentration So that's relating on the fact that how you manage to reduce the aviation without affecting your effluent ammonia We didn't necessarily reduce aeration but we just changed the delivery of dissolved oxygen delivery so the set point So what we found was that the intermittent and the ramping up and the ramping down of dissolved oxygen actually contributed to nitrous oxide The higher DO contributed through the hydroxyl amine oxidation pathway and also we found under low DO we're getting nitrite accumulation so nitrify denitrification So it was about providing I guess a more even supply of oxygen not necessarily reducing it down but also we were I guess there was evidence of I guess simultaneous nitrification denitrification occurring as well at those levels But yeah I just want to stress the fact that it's not so much necessarily reducing we went from an intermittent aeration profile to more of an even supply of oxygen Does it have an effect on the energy consumption of the plant? I guess that what we found is we're actually at times over aerating the plant so we're actually over aerating and as a result we're actually getting a lot of nitrate accumulating as well So by reducing that waste we're able to achieve a reduction not only in nitrous oxide but prevented nitrate from accumulating and also minimized energy consumption that was associated with the over aeration There are many many questions in the chat maybe you can answer some of them directly specific questions there was a question on the model the University of Queensland youth What I can also do is provide the links to the papers all this work was published particularly around the model as well Yeah I think most of the questions can be found in the papers and well just to finalize how did the biology respond to operating at continuous lower DO level? In terms of the biomass concentration and the settlability yeah so there was a small I guess the sludge volume index suffered a little bit but not too much so there was an increase in a reduction in the settlability but it was still manageable Thanks a lot Ben I think we need to move to the next and last speaker Thank you Dr. Basilia Basilaki She's a research fellow at the Brunel University in London and is working on process emissions and on the development of artificial intelligence wastewater and environmental solutions And she's author of seven papers on this topic including a summary of 10 years of monitoring of N2O which was the first time a global data set was developed and analyzed across process types so I'm sure most of the audience are familiar with this work She's also the first author of chapter 6 of the book which focuses on full scale emission results and also is a author of chapter 10 that is focused on knowledge based data that has driven approaches for assessing G&G emissions from wastewater treatment plants And in her presentation she will give us a summary of these two chapters Basilia, the floor is yours, thank you Thank you, thank you Maite and thank you everyone for being here today just to check that I can control the presentation Okay Just a little bit background about our R&D activities at Brunel University We are a group of 20 researchers working with Professor Evina Katsu on understanding and mitigation of process emissions and on tools and frameworks for the sustainability and circularity assessment and monitoring control and optimization of water processing In this presentation we will discuss about chapter 6 and 10 of the book on process emissions and specifically about global N2O studies and innovative data use There have been more than 20 years since the scientific community started monitoring the N2O emissions in full scale reactors Today we will summarize what we have learned until now from all these studies and how we can use the knowledge accumulated to improve our understanding on process emissions notification We now know that the significant amount of N2O can be generated and emitted during biological nutrients removal and two emissions can contribute up to 78% to the operational carbon footprint of the plants This means that in order to work towards carbon neutral plants and two emissions and termifications need to be considered This is not very straightforward since studies have resulted in conflicting findings in terms of triggering operational conditions and long term dynamics cannot be fully explained yet This is because the complex relationships as we saw even in the previous presentations that exist between the N2O generation and operational conditions and because different interactions of the operational variables can trigger a different response on the N2O generation But now let's see what we report in this new publication on process emissions We have analyzed the quantified emission factors, triggering mechanisms and mitigation measures in over 80 full scale wastewater treatment systems The taxonomy reported in the book considers the process type and wastewater treatment plant size, environmental and operational conditions process control parameters and efficiency indicators We have also proposed mitigation measures for the reduction of high end tourists and innovative data utilization methods based on the existing studies To be able to better understand the results and the work undertaken, let's first see some typical reviles of nitrous eyes As you can see in the figure on the top left, there is a huge seasonal variation This is an example from a wastewater treatment plant in Netherlands and we can see that along the year emissions can be from almost zero to up to 200 kilograms of end to over day Studies have also shown that it is possible for emissions to vary from one year to the next This means that the long term campaigns are necessary to quantify emission factors and the telemetrication strategy needs to be dynamic to cope with the dynamic variation of triggering factors We also expect a deunal variation, as you can see in the typical file on the figure on the right, and as we saw in the previous presentation, spatial variation One of the first things we did when we collected all this data from the studies is that we developed workflows of the emission factors, as you can see in the figure on the top left for different configurations of processes, specific processes such as sidestream processes treating high strength streams are associated with high risk of elevated emissions While in conventional processes, we expect the emissions to range between 0.2 to 2% of the nitrogen load and this is a huge variation in terms of carbon footprint impacts But are all these emission factors reliable? In the figure on the bottom, you can see the emission factor values with respect to the length of the monitoring period The average emission factor for long term campaigns is equal to 1.7% of the nitrogen load, whereas monitoring campaigns lasting less than a month have an average emission factor equal to 0.7% And this is exactly because we saw that emissions can vary significantly during the year and this can impact a lot the reliability of the emission factors we have quantified Specific patterns of end-to-emissions have been also identified, they can depend on the process performance and control the type of biological treatment, the sludge treatment among other factors that you can find in the relevant chapter in the book As an example, plants that have anaerobic digestion on site and feed the anaerobic supernatant into the main line into the biological process have a median emission factor equal to 1.5% of the nitrogen load On the other hand wastewater treatment plants that have applied sludge treatment strategies such as sludge dye have a median emission factor equal to 0.11% of the nitrogen load And what does this mean? It means that we should do better in the future, not only in the design of the monitoring strategies but also in the data we collect and report to develop meaningful benchmarks for different processes In the book we have also summarized the knowledge accumulated for triggering conditions for specific groups of processes and the respected mitigation measures that have been identified This is a simplified example for the processes removing nitrogen, several conditions have been linked with elevated emissions risks such as insufficient dissolved oxygen We have seen increased stripping due to elevated aeration rate and elevated nitrite concentration among several other factors The knowledge on what triggers emissions has resulted in the development of mitigation strategies related to the control of ammonia oxidation rates, the modification of the control of dissolved oxygen improved understanding of the effect of the mixed liquor suspended soil concentration and several other mitigation factors as we saw in the slides in the previous presentations We have also seen that the role of data in understanding and mitigation of emissions is very very significant In fact several studies have shown that the sensor data and laboratory analysis from wastewater treatment processes contain hidden information that can be valorized to explain and control the long term dynamics of end to emissions and triggering operational conditions We believe that there are definitely better ways to monitor, control and manage our wastewater treatment processes So in the book we present some recent success stories and practical examples of the capabilities opening up when we leverage the information hidden in raw data I will present some examples today from chapter 10 of the book, this is an example from Cobaltwater Global who have developed an end to risk decision support system The tool uses knowledge based artificial intelligence and machine learning to propose process adjustment Here we can see an example with the proposed dissolved oxygen in light blue and the actual dissolved oxygen in grey Before and after the adjustment we can see when the adjustment starts because the actual dissolved oxygen is much closer to the suggested DO Here we can see the behavior of nitrous oxide before and after the adjustment, it is the line in purple We can see that after the adjustment the end to all it uses significantly In light green we can see a machine learning model that predicted this end to all reduction In this example up to 70% reduction of end to all was achieved Several examples of data utilization from our own work and other works are presented in the book We can see how we can use data to minimize end to all sampling requirements for the reliable quantification of annual emission factors This is when we do not want to monitor for one year continuously In practice we have shown that AI guided sampling strategies can result in the most reliable annual end to emission factors compared to conventional intermediate sampling methods such as monthly sampling We also see examples on how to detect abnormal events, change points in the behavior of the system and how to apply clustering classification regression algorithms to translate data into actionable information link the end to all ranges with specific operational conditions, predict the range of emissions based on the operational environmental conditions and provide feedback for the mitigation measures The introduction of data driven models allows water professionals to understand and improve the environmental performance profile of their operations and potentially dynamically control the emissions So to conclude just a short take home message, process based and tool quantification, benchmarking and mitigation we know it is still challenging but structured approaches for monitoring and knowledge discovery from wastewater databases reinforced by a combination of domain knowledge, data mining techniques and mechanistic models are powerful tools to support wastewater treatment plant carbon neutrality targets and to facilitate the integration of sustainability metrics in the decision making I would like to thank a lot to also the co-authors of Chapter 6, Maite, Jaoran and Tevina and also for Chapter 10, Jose Poro with the lead author and Giacomo Velanti and Tevina Thanks a lot Thank you very much, Lozalea That was the last talk And very interesting, we have as well several questions for you, maybe just a quick one and then we can go to the five minutes, all of us, Kasia One is why does the mission go up during part of the year that you have shown? There are several interpretations for this For example, we've seen that in several plants there is a peak of emissions and increase of emissions during Easter so when the temperature goes from lower temperatures and increases and this has been potentially attributed to inhibition of nitrite, of NOB, nitrite oxidizing bacteria but there are several, several environmental and operational factors that can lead to temporal changes in the emissions behavior Thank you Now I will invite all of the panelists if they can switch on the microphones and the cameras just to have maybe five to six minutes of discussion We can address some of the comments we have and questions we have been seeing in the Q&A section Okay, it's one maybe more general that what will be the best method to easily monitor and to all Maybe now this one to Vanessa if you can comment It will depend on the type of plan, you think maybe a plan-wide model will benefit some plants and others can benefit more from process units and also it will depend I guess on the how much they can spend on the monitoring Of course there are several factors that need to be considered Sorry Well as I said in the presentation first of all I think you have to ask yourself what do you want to measure So what is the goal of your measurements? I mean if you just need the total amount of and to omit it Just to have an first idea on how high it will be and maybe have to comply with some greenhouse gas emissions protocols and so on Then of course you can go for the overall approach If you need more detailed information because you want to link the entry emissions with the process condition because you want to implement some mitigation strategies at the plant And of course this overall emission measurements will not give you this kind of degree of information So you have to then go for a unit, a process unit measurement I can go more into details Of course there are also some local conditions that needs to be considered I think it was already mentioned if you have for example surface aerators and not bubble aerators Then of course you have to consider this when deciding to go for the hood or to have the measurement in the liquid phase And as you said the cost is also a very important point especially when you are We are designing also the duration of the campaign And I mean of course if it's possible to have a very long measurement campaign it's always an advantage But sometimes it's not possible due to the quite high input of resource in all of this And I think the most important thing is that you have of course to catch some seasonal variation Probably you know your plant best, so if you know that there are seasonal variation in the temperature But also in the loading then of course to have measurement in both these periods Of course you have to have at least some weeks, some months would be even better Really to monitor also the fluctuation of the load and how they are impacting the plant And the entry emissions And yeah so in the end I think it's always, it has to be really specifically defined First of all for the goal of the measurement and also to consider the site conditions you have at the plant Just a very generally set Following that is interesting, the years have been becoming more and more popular to monitor this oven tool You can also plug these sensors in the sky and get this data almost online And one of the audience and person in the audience was asking if it would be necessary to develop this two phase model For predicting N2O since we have most of the data from the liquid phase Monitor at full scale and maybe that's a question for how we run Do you think it's, yeah, we could target to develop this model and also including this stripping It seems quite critical to stimulate emissions of N2O Yeah, I see where that question is coming from So liquid monitoring of N2O is obviously more easily available rather than having a complicated hood and analyzer setup The two phase model I'm not sure I'll completely get you But it's sort of in all the models we model until generations in liquid phase Then we model a transfer from liquid to gas phase So that is a two phase model and I think maybe the people who asked this was referring to a simple mass transfer model So then the question is really how do you estimate the KLA The model itself is just a simple equation The real issue is how do you determine KLA And I think I saw in the chat box Michael was there who was the CTO of Unicense We just had a chat some days ago about the KLA estimation It is quite tricky Many treatment plants is doing having a hood setup as well as a liquid N2O monitoring simultaneously So that they can calibrate the KLA with the hood measurement and the liquid management measurement Then they determine the relationship and use such equation to estimate N2O emissions But then for treatment plants who do not have hood setup things became a little bit tricky So the carbon recommended method by Unicense is a method developed by Jeff Foley at the University of Queensland He used small reactors and also grabbed samples at full scale He used an empirical equation to estimate the KLA which may be good but may also have some issues require some improvement To summarize it will be the most accurate will be to use an experimental KLA approach to have this hood there And experimented it test the KLA in your plant If you don't have the resources then you can go to this more maybe less accurate But better than IPCC And maybe a final question there are many but just to finish this webinar maybe this one goes on Ben What are the main motivations for which water treatment plant managers to reduce N2O? Is it claiming emissions reductions? Is it improving process performance? Is it to help reduce global warming? That's a very good question. Our initial driver was that Australia introduced a carbon tax And we had the choice of using a generic emission factor or going out and measure the emissions So the use of a generic emission factor we didn't know whether our emissions were above or below it So what we did was we took the chance to go out and measure directly And what we found was that our emissions were much higher than the use of a generic emission factor Which meant that our liability in carbon tax was going to be much higher as well What happened is quite quickly that there was a change in government and the carbon tax was revoked But we then had all this information in terms about what our true emissions were And then trying I guess to identify steps that we could reduce their emissions And I guess also so there was no I guess no financial incentives incentive It was just I guess an opportunity working with the University of Queensland We'd done this work and then we thought okay what can we do next? How can we reduce these emissions? And basically there was no real driver There was no real incentive at the time. It was just I guess an opportunity to work Industry working with the university collaboratively But now the driver has returned. We've set a target of net zero by 2030 And so we now need to go out and repeat some of this work and monitor again To see if our emissions have changed So yeah initially there was no real driver But now that driver has returned and there will be a financial incentive to continue this work Well thank you very much. I think that we are running out of time Maybe well first of all to thank all the audience I hope you find this webinar interesting for yeah it was a thing We covered most of the current state of the art of the N2O emissions and monitoring And we saw some examples of how mitigation strategies can be applied full scale And now you will just just to finish with some yeah Announcement on next webinar This one that will take place on the 25th of May on complete ammonia oxidizers Also I would like to mention that there is another master class of this webinar series on the 23rd of June That will be focused just on methane emissions For you if you want to start registering And next slide please We have another announcement I hope to see many of you in the upcoming IwoWater Congress That has to be had to be postponed this year and this year will take place in the fantastic Copenhagen in September And yeah I think yeah and also please if you're not part of IWA You can you can join this 20% discount on the new memberships And yes thank you all the panelists And thank you IWA to organize these fantastic webinars And thank you all for your attendance and I hope you'll see you soon in the future Thank you Thank you Mike T Thank you Thank you very much