NAXOS 2018 Assessment of wastewater N2O generation using - - PowerPoint PPT Presentation

▶
naxos 2018 assessment of wastewater n2o generation using
SMART_READER_LITE
LIVE PREVIEW

NAXOS 2018 Assessment of wastewater N2O generation using - - PowerPoint PPT Presentation

Aug 21, 2023 186 likes •458 views

NAXOS 2018 Assessment of wastewater N2O generation using multivariate techniques Vasilaki V. 1 , ConcaV. 3 , Frison N. 3 , Mousavi A. 2 , Fatone F. 4 , Katsou E. 1 15/5/2018 Overview 21 June 2018 Introduction SMARTPlant Aim and

slide-1
SLIDE 1

15/5/2018

Assessment of wastewater N2O generation using multivariate techniques

Vasilaki V.1, ConcaV.3, Frison N.3, Mousavi A.2, Fatone F.4, Katsou E.1

NAXOS 2018

slide-2
SLIDE 2

Brunel University London 21 June 2018

2

Introduction SMART‐Plant Aim and objectives Methodology Results

  • DBSCAN
  • Hierarchical clustering
  • SAX

 Conclusions

Overview

slide-3
SLIDE 3

Brunel University London Brunel University London Chemical Intermediates (VFA, N,P derivatives) Struvite P‐rich compost Biocomposites from Cellulose & PHA Biogas Water reuse

Agriculture Chemical Industry Construction Industry Multipurpose water reuse Water‐energy nexus

Biofuel from cellulosic sludge

WWTP

Scale-up of low-carbon footprint MAterial Recovery Techniques for upgrading existing wastewater treatment Plants (SMART-Plant)

Sensor Networks Smart Water Management Data Management Modelling and

  • ptimization
slide-4
SLIDE 4

Brunel University London

I NTCATCH Ad-Bio C-FOOT-CTRL

Research Projects Research Activities

Brunel University London

Scale-up of low-carbon footprint MAterial Recovery Techniques for upgrading existing wastewater treatment Plants (SMART-Plant)

Urban W astew ater W ater P-Com post Cellulose Energy Struvite

Reduce

Energy GHGs

Biopolym ers

Online Energy and GHG monitoring Environmental benefits Economic benefits

Sm artPlant

slide-5
SLIDE 5

Brunel University London

I NTCATCH Ad-Bio C-FOOT-CTRL

Research Projects Research Activities

Brunel University London

Scale-up of low-carbon footprint MAterial Recovery Techniques for upgrading existing wastewater treatment Plants (SMART-Plant)

Data Acquisition Outliers detection Pattern recognition Dependencies identification Data analysis CF-EF assessment tool Identification of set-points that

  • ptimize carbon footprint

Identification of patterns and dependencies

Equalization tank for anaerobic supernatant BACS storage Date‐Time Level (m) Level (m) Conductivity (µS/cm) ORP (mV) DO (mg/L) MLSS (mg/L) pH N2O (μM) MLSSin (mg/L) MLSSout (mg/L) Flow-ratein (m3/h) Flow-rateout (m3/h) polyelectrolyte dosage Level (m) pH Flow-ratein sludge (m3/h) Flow-ratein polyelectrolyte (m3/h) Level (m) Level (m) Short‐cut Sequencing Batch Reactor (scSBR) Dynamic thickening Fermentation unit (SBFR) Solid/Liquid separation and the carbon source storage tank (screw press) Enhanced Primary Separation of municipal wastewater unit Crystallizer Date‐Time Level (m) Level (m) pH Temperature (°C) Pressure (bar) Level (m) Flow‐rate recycle (m3/h) Flow‐rate Permeate (m3/h) pH Level (m) Conductivity (µS/cm) DO (mg/L) pH N2O (μM) Level (m) Conductivity (µS/cm) DO (mg/L) pH N2O (μM) PHA‐accumulating biomass selection SBR+U:V Sequencing Batch Fermentation Reactor Ultrafiltration unit Nitritation SBR PHA‐accumulating biomass selection SBR Cl N2O C NH4- N PF NO3- N PF Influent NH4-N C NO3-N C DO1 DO2 DO3 NO2-N Temp 1 1 0.01 10.38 1.50 4593 1.26 5.18 0.47 0.71 1.57 15.6 2 0.01 7.03 3.26 3559 0.69 3.86 0.01 0.30 0.99 3 0.1 15.00 1.50 3849 1.63 9.16 1.06 0.76 1.73 4 0.06 16.45 0.27 8665 9.77 4.79 1.36 0.68 1.55 2 5 0.73 15.05 1.92 3978 1.57 8.22 0.96 1.47 2.21 11.3 6 0.24 9.18 4.09 2994 0.71 5.71 0.02 0.59 1.34 7 0.19 11.21 0.90 11506 4.03 5.05 1.86 2.22 2.10 3 8 2.53 15.55 4.33 3111 1.23 11.33 0.65 2.42 2.39 11.6 9 1.18 10.22 3.34 3235 0.71 6.07 0.02 0.49 1.42 10 2.39 17.24 1.01 3539 1.78 6.73 0.98 1.25 2.25 11 1.72 21.99 0.11 9669 9.66 4.13 2.02 0.07 2.08 4 12 5.25 17.11 0.24 3422 1.31 3.87 0.85 2.37 2.27 11.6 13 3.39 11.42 0.65 3040 0.77 2.45 0.00 1.38 1.42 14 3.96 23.25 0.11 7489 9.56 3.10 1.96 1.74 1.96 15 0.96 8.62 0.14 9824 0.97 1.48 0.92 2.07 1.92 5 16 7.10 19.37 0.85 3322 2.01 6.52 1.66 1.57 1.93 3.98 16.2 17 3.39 11.16 2.48 2665 0.90 3.81 0.06 0.71 1.63 0.77 18 2.12 8.29 6.08 2335 0.85 11.44 1.08 2.24 2.26 1.28

Sm artPlant

slide-6
SLIDE 6

Brunel University London 6/21/2018

6

Aim and objectives

Investigation N2O accumulation profiles and dependencies with

  • perating parameters in a full‐scale reactor

Implementation of appropriate statistical methods to identify

  • perating conditions for mitigating N2O generation

Motif and pattern analysis of critical parameters monitored online and investigation of their effect on N2O generation in the bioreactor

slide-7
SLIDE 7

Brunel University London 21 June 2018 Presentation Title

7

Schreiber bioreactor simultaneous nitrification and denitrification process by a time‐based intermittent aeration. The excess phosphorus chemically removed poly‐alluminium‐chloride (PAC)

Methodology

Air Influent Diffusers Rotating diffuser support Return activated sludge Effluent Secondary clarifiers 4571 m3 PE: 40000

Flow‐chart

slide-8
SLIDE 8

Brunel University London 21 June 2018 Presentation Title

8

Methodology – N2O monitoring

Clark-type electrode (Unisense, Aarhus, Denmark). Dissolved N2O

Online Offline DO COD influent NH4-N effluent

ORP

TN influent NO3-N effluent TSS TP influent NO2-N effluent N2O COD/Ntot TP influent Flow-rate COD effluent pH effluent Blowers flow-rate TN efffluent

slide-9
SLIDE 9

Brunel University London 21 June 2018 Presentation Title

9

Methodology - analysis

Anomalous event detection

 Uncommon diurnal patterns Sudden changes in the underlying system  Sensor faults

Density‐Based Spatial Clustering of Applications with Noise – DBSCAN (1)

 Neighborhood distance epsilon (eps) : The radius of the neighborhoods around a data point p.  Effective method with medium sized data sets (3).  Widely used anomaly detection problems (2)  minPts: The minimum number of data points required in a neighborhood to define a cluster.

slide-10
SLIDE 10

Brunel University London

 Identify groups of similar days online operating variables

Hierarchical clustering (4).

 Agglomerative: clusters incrementally, producing a dendrogram

 Ward's method (5)

Motif discovery – Symbolic Aggregate approXimation (SAX) (6)

 Repeated motifs in a time series

Methodology - analysis

*Data standardization

 Convert 24h hourly series to piecewise aggregate approximation (PAA) representation Convert the PAA into a string of symbols

adfefdaa

slide-11
SLIDE 11

Brunel University London 21 June 2018 Presentation Title

11

Results

Nitrification efficiency: ~80% and Denitrification efficiency: ~75%.

Variables Values Offline variables Average Std COD influent (mg COD/ L) 234.6 88.5 TN influent (mg/L) 24.6 3.3 TP influent (mg/ L) 4.8 0.9 COD effluent (mg/ L) 26.8 3.7 TN efffluent (mg/ L) 6.4 `1 TP influent (mg/ L) 1.6 0.7 pH effluent 7.1 0.1

Mean: 9.4

0,0 2,0 4,0 6,0 8,0 10,0 12,0 14,0 16,0 18,0 12/7/2017 27/7/2017 11/8/2017 26/8/2017 10/9/2017 25/9/2017

COD/TN Date

Lowest COD/N

0,6 0,65 0,7 0,75 0,8 0,85 0,9 0,95 1 0,05 0,1

Nitification efficiency (%)

N2O (mg/L) 0,0 2,0 4,0 6,0 8,0 10,0 12,0 14,0 16,0 18,0 0,05 0,1 0,15

COD/TN N2O (mg/L) 0,05 0,1 0,15 0,2 0,25 0,3 0,35 0,4 13/7/2017 2/8/2017 22/8/2017 11/9/2017 N2O (mg/L) Monitoring period (h)

N2O accumulation profile

slide-12
SLIDE 12

Brunel University London

  • 120
  • 100
  • 80
  • 60
  • 40
  • 20

20 0,01 0,02 0,03 0,04 0,05 0,06 0,07 25 35 45 55 65 75

ORP (mV) N2O (mg/L)-

Operating time (h) N2O Emissions ORP

0,01 0,02 0,03 0,04 0,05 0,06 0,07

  • 50
  • 40
  • 30
  • 20
  • 10

10 20 30 40 25 35 45 55 65 75

N2O (mg/L) ORP slope

Operating time (h) ORP slope N2O Emissions

21 June 2018 Presentation Title

12

Results

N2O accumulation ORP and DO

Strong connection is

  • indicated. However, is

this relationship consistent in the system? Under which conditions exists? Spearman’s correlation: DO, N2O 0.62 ORP, N2O 0.72 ORP, log(DO) 0.80

0,01 0,02 0,03 0,04 0,05 0,06 0,07 0,05 0,1 0,15 0,2 0,25 0,3 25 35 45 55 65 75

N2O (mg/L DO (mg/L)

Operating time (h) do N2O Emissions

slide-13
SLIDE 13

Brunel University London 21 June 2018

13

Results

Outliers detection

Each point: d-dimensions, where d=24 is the number of elements in the daily time series. Red points: outliers Influent flow-rate DO ORP

Outliers Influent DO ORP TSS Days 7 2 5 1 Sensor errors 1

slide-14
SLIDE 14

Brunel University London 21 June 2018

14

Results

Outliers detection

Outliers Influent DO ORP TSS Days 7 2 5 1 Sensor errors 1

Sensor fault Unusual behavior

slide-15
SLIDE 15

Brunel University London 21 June 2018

15

Results

Outliers detection

Linked with precipitation events Unusual diurnal profile of ORP linked with high influent flow-rates N2O accumulation = 0

slide-16
SLIDE 16

Brunel University London 21 June 2018 Presentation Title

16

Results

Cluster Cluster

Mean (mg/L) 0.19 0.23 0.29 Mean (mg/L) ‐32 ‐2

i) specific ranges of DO and ORP N2O accumulation and ii) examine the effect of other parameters (monitored online and

  • ffline) that could impact the N2O measurements in periods

with similar ranges of ORP and DO.

Group Cluster ORP Cluster DO Days 1 1 1 19 2 1 2 14 3 1 3 1 4 2 2 3 5 2 3 15 Disturbances

  • 10

i) variations impacted the N2O accumulation ii) specific ranges of DO and ORP high N2O accumulation iii) examine the effect of other parameters with steady profiles of ORP and DO.

slide-17
SLIDE 17

Brunel University London 21 June 2018 Presentation Title

17

Results

Clusters with N2O accumulation DO sensor fault Influent and ORP uncommon patters ORP uncommon pattern

Group Cluster ORP Cluster DO Days 1 1 1 19 2 1 2 14 3 1 3 1 4 2 2 3 5 2 3 15

slide-18
SLIDE 18

Brunel University London 21 June 2018 Presentation Title

18

Results

Clusters with N2O accumulation

Group Cluster ORP Cluster DO Days 1 1 1 19 2 1 2 14 3 1 3 1 4 2 2 3 5 2 3 15

Cluster 1: Average accumulation of N2O = ~ 0.008 mg/L N2O peak: < 0.09 mg/L until 31st July 31st July: very high COD and TN concentrations COD and TN peaks

slide-19
SLIDE 19

Brunel University London 21 June 2018 Presentation Title

19

Results

Clusters with N2O emissions

Group Cluster ORP Cluster DO Days 1 1 1 19 2 1 2 14 3 1 3 1 4 2 2 3 5 2 3 15

Cluster 2: Average accumulation of N2O = ~ 0.02 mg/L N2O peak: > 0.1 mg/L most days

COD load = 1300 kg/day TSS peak TSS peak

slide-20
SLIDE 20

Brunel University London 21 June 2018 Presentation Title

20

Results

Clusters with N2O emissions

Group Cluster ORP Cluster DO Days 1 1 1 19 2 1 2 14 3 1 3 1 4 2 2 3 5 2 3 15

Cluster 5: Average accumulation of N2O = ~ 0.08 mg/L N2O peak: > 0.15 mg/L most days

slide-21
SLIDE 21

Brunel University London 21 June 2018 Presentation Title

21

Results

Motif investigation ORP (3) /DO (6) Shape-based Clustering

N2O accumulates mainly during denitrification

‐60 ‐50 ‐40 ‐30 ‐20 ‐10 10 20 30

0,1 0,2 0,3 0,4 0,5 1 3 5 7 9 11 13 15 17 19 21 23 25 27 29 31 33 35 37 39 41 43 45 47 ORP (mV) DO (mg/L) Hours (h) DO ORP

‐60 ‐50 ‐40 ‐30 ‐20 ‐10 10 20 30

0,05 0,1 0,15 0,2 0,25 0,3 1 3 5 7 9 11 13 15 17 19 21 23 25 27 29 31 33 35 37 39 41 43 45 47 ORP (mV) N2O (mg/L) Hours (h) N2O ORP

slide-22
SLIDE 22

Brunel University London ‐50 ‐40 ‐30 ‐20 ‐10 10 20 30

0,05 0,1 0,15 0,2 0,25 0,3 0,35 0,4 0,45 1 3 5 7 9 11 13 15 17 19 21 23 25 27 29 31 33 35 37 39 41 43 45 47 ORP (mV) DO (mg/L) Hours (h) DO ORP

21 June 2018 Presentation Title

22

Results

‐50 ‐40 ‐30 ‐20 ‐10 10 20 30

0,05 0,1 0,15 0,2 0,25 0,3 0,35 1 3 5 7 9 11 13 15 17 19 21 23 25 27 29 31 33 35 37 39 41 43 45 47 ORP (mV) N2O(mg/L) Hours (h) N2O ORP

Motif investigation ORP (1) /DO (7) Shape-based Clustering

N2O accumulates mainly during nitrification

slide-23
SLIDE 23

Brunel University London ‐20 ‐10 10 20 30 40

0,1 0,2 0,3 0,4 0,5 0,6 1 3 5 7 9 11 13 15 17 19 21 23 25 27 29 31 33 35 37 39 41 43 45 47 ORP (mV) DO (mg/L) Hours (h) DO ORP

21 June 2018 Presentation Title

23

Results

‐20 ‐10 10 20 30 40

0,05 0,1 0,15 0,2 0,25 0,3 1 3 5 7 9 11 13 15 17 19 21 23 25 27 29 31 33 35 37 39 41 43 45 47 ORP (mV) N2O(mg/L) Hours (h) N2O ORP

Motif investigation ORP (2) /DO (2) Shape-based Clustering

N2O accumulates mainly during denitrification

slide-24
SLIDE 24

Brunel University London 21 June 2018 Presentation Title

24

Conclusions

 Significant correlation between N2O accumulation, DO and ORP equal to 0.62 and 0.72 respectively  No significant relationship between N2O accumulation, COD/TN and NH4-N removal efficiency  DBSCAN Detect days with sensor faults and unusual patterns in the parameters monitored online  High influent flow-rates (peak>850 m3/h) in the system coincided with unusual patterns of ORP  Hierarchical clustering 5 clusters with different diurnal ranges of ORP and DO. Clusters with higher ORP and DO peaks mainly linked with higher N2O accumulation  SAX identify similar daily motifs of ORP and DO, coinciding with different N2O accumulation behavior in the system

slide-25
SLIDE 25

Brunel University London 21 June 2018

25

References

(1) Ester, M., Kriegel, H.P., Sander, J. and Xu, X., 1996, August. A density‐based algorithm for discovering clusters in large spatial databases with noise. In Kdd (Vol. 96, No. 34, pp. 226‐ 231). (2) Çelik, M., Dadaşer‐Çelik, F. and Dokuz, A.Ş., 2011, June. Anomaly detection in temperature data using dbscan algorithm. In Innovations in Intelligent Systems and Applications (INISTA), 2011 International Symposium on (pp. 91‐95). IEEE. (3) Kaufman, L.R. and Rousseeuw, P., PJ (1990) Finding groups in data: An introduction to cluster analysis. Hoboken NJ John Wiley & Sons Inc, 725. (4) Ward Jr, J.H., 1963. Hierarchical grouping to optimize an objective function. J. Am. Stat.

  • Assoc. 58, 236–244.

(5) Lin, J., Keogh, E., Wei, L., & Lonardi, S. (2007). Experiencing SAX: a novel symbolic representation of time series. Data Mining and knowledge discovery, 15(2), 107‐144.

Acknowledgements This paper is supported by the Horizon 2020 research and innovation programme, SMART‐Plant under grant agreement No 690323. The authors would like to acknowledge the Royal Society for the funding of the current research:Ad‐Bio: Advanced Biological Wastewater Treatment Processes, Newton Advanced Fellowship2015/R2.

slide-26
SLIDE 26

Thank you for your attention!