Photoelectrochemical Chemical Oxygen Demand …®-COD...Civil and Resource Engineering Dalhousie University Photoelectrochemical Chemical Oxygen Demand Analysis in Drinking Water AminaStoddart

Civil and Resource Engineering Dalhousie University

Photoelectrochemical Chemical Oxygen Demand Analysis in Drinking Water Amina Stoddart

February 11, 2016

Introduction

•  Natural organic matter (NOM) is a critical target for drinking water treatment

•  NOM can be associated with – Taste, odour, colour issues – Coagulant, oxidant demand – DBP precursors

– We have a number of tools for bulk NOM estimation: DOC, TOC, UV254, SUVA

Chemical Oxygen Demand (COD) Measurement in Drinking Water

•  Traditional NOM surrogates may not be suitable for assessing NOM removal in all cases – UV254, SUVA

•  Rely on aromaticity, which is not a chemical feature of many organic compounds, example sugars

– Carbon (e.g., as TOC, DOC) •  Does not quantify the reactivity of the organic

What is Chemical Oxygen Demand?

+ O2 à CO2 + H2O + NH3

TOC measures conversion to CO2

What is Chemical Oxygen Demand?

+ O2 à CO2 + H2O + NH3

COD measures “demand” for oxygen

TOC measures conversion to CO2

Why is COD not often used in Drinking Water?

•  The traditional method for COD determination is to oxidize with potassium dichromate under acidic conditions

•  Issues: – Sensitivity – Use of hazardous chemicals

•  Dichromate, mercury, surfuric acid

– Analysis time •  Hours

Photoelectrochemical COD (peCOD) Analysis •  Safe for operator

–  No hazardous chemicals –  Single reagent (electrolyte)

•  Takes 5-10 min –  Can automate –  Potential for online

measurement •  Low range

–  MDL = 0.5 mg/L (using modified procedure)

•  Uses green chemistry –  No hazardous wastes

Working Principle: peCOD

Technical Approach

1.  Conducted initial method validation with model organic compounds a.  Compared peCOD of carboxylic acids, amino acids and

reference compounds to the calculated theoretical oxygen demand (ThOD)

b.  Verified peCOD applicability in the drinking water NOM range of concern

2.  Tested technology at various drinking water treatment plants

3.  Monitored full-scale drinking water biofiltration

Method Validation: Comparison of peCOD and ThOD for Amino Acids

Figure: Stoddart, A. K., & Gagnon, G. A. (2014). Application of photoelectrochemical chemical oxygen demand to drinking water. Journal: American Water Works Association, 106(9).

y = 1.14x - 1.18 R² = 0.98

y = 1.01x - 0.62 R² = 0.97

y = 1.01x - 0.87 R² = 0.99

0.0

5.0

10.0

15.0

20.0

0.0 2.0 4.0 6.0 8.0 10.0 12.0 14.0 16.0

peC

OD

– m

g/L

ThOD – mg/L

Phenylalanine

Tyrosine

Tryptophan

Method Validation: Comparison of peCOD and ThOD for Amino Acids


y = 1.14x - 1.18 R² = 0.98

y = 1.01x - 0.62 R² = 0.97

y = 1.01x - 0.87 R² = 0.99

0.0

5.0

10.0

15.0

20.0

0.0 2.0 4.0 6.0 8.0 10.0 12.0 14.0 16.0

peC

OD

– m

g/L

ThOD – mg/L

Phenylalanine

Tyrosine

Tryptophan

Slope value of unity would demonstrate that peCOD was a complete

predictor of ThOD

Method Validation: Comparison of peCOD and ThOD Carboxylic Acids


y = 1.47x - 0.13 R² = 0.99

y = 0.96x - 0.58 R² = 0.98

y = 0.79x - 1.34 R² = 0.94

0.0

5.0

10.0

0.0 2.0 4.0 6.0 8.0 10.0 12.0 14.0 16.0

peC

OD

– m

g/L

ThOD – mg/L

Na-Oxalate

Na-Formate

Na-Acetate

Method Validation: Comparison of peCOD and ThOD Carboxylic Acids


y = 1.47x - 0.13 R² = 0.99

y = 0.96x - 0.58 R² = 0.98

y = 0.79x - 1.34 R² = 0.94

0.0

5.0

10.0

0.0 2.0 4.0 6.0 8.0 10.0 12.0 14.0 16.0

peC

OD

– m

g/L

ThOD – mg/L

Na-Oxalate

Na-Formate

Na-Acetate

Slope value of unity would demonstrate that peCOD was a complete

predictor of ThOD

Method Validation: Comparison of peCOD and TOC

•  peCOD detectable at TOC concentrations characteristic of raw and treated water –  i.e., 1-5 mg C/L

•  peCOD:TOC ratios were predictable based on stoichiometry of the oxidation reaction –  i.e., oxygen to carbon

ratio

y = 3.24x - 1.52 R² = 0.99

y = 3.05x - 1.10 R² = 0.96

y = 3.11x - 1.30 R² = 0.99

0.0

5.0

10.0

15.0

20.0

0.0 2.0 4.0 6.0

peC

OD

– m

g/L

TOC – mg/L

Phenylalanine

Tyrosine

Tryptophan


Method Validation: Comparison of peCOD and TOC

•  peCOD detectable at TOC concentrations characteristic of raw and treated water –  i.e., 1-5 mg C/L

•  peCOD:TOC ratios were predictable based on stoichiometry of the oxidation reaction –  i.e., oxygen to carbon

ratio


y = 1.03x - 0.34 R² = 0.99

y = 1.33x - 0.88 R² = 0.98

y = 2.14x - 1.62 R² = 0.94

0.0

5.0

10.0

0.0 2.0 4.0 6.0

peC

OD

– m

g/L

TOC – mg/L

Na-Oxalate

Na-Formate

Na-Acetate

Method Validation: Various Treatment Plants

Figures adapted from: Stoddart, A. K., & Gagnon, G. A. (2014). Application of photoelectrochemical chemical oxygen demand to drinking water. Journal: American Water Works Association, 106(9).

0.0

1.0

2.0

3.0

4.0

5.0

6.0

7.0

8.0

Con

cent

ratio

n –

mg/

L

peCOD TOC DOC

0.0

5.0

10.0

15.0

20.0

25.0

Con

cent

ratio

n –

mg/

L

peCOD TOC DOC

Direct Biofiltration Plant Conventional Filtration Plant

Method Validation: Various Treatment Plants

Figures adapted from: Stoddart, A. K., & Gagnon, G. A. (2014). Application of photoelectrochemical chemical oxygen demand to drinking water. Journal: American Water Works Association, 106(9).

Membrane Treatment Plant Conventional Filtration Plant

0.0 2.0 4.0 6.0 8.0

10.0 12.0 14.0 16.0 18.0 20.0

Con

cent

ratio

n –

mg/

L

peCOD TOC DOC

0.0 2.0 4.0 6.0 8.0

10.0 12.0 14.0 16.0 18.0 20.0

Con

cent

ratio

n –

mg/

L

peCOD TOC DOC

Method Validation: Various Treatment Plants in Nova Scotia - peCOD and TOC


y = 2.69x - 0.49 R² = 0.64

0.0

5.0

10.0

15.0

20.0

25.0

0.0 1.0 2.0 3.0 4.0 5.0 6.0 7.0

peC

OD

– m

g/L

TOC – mg/L

Bennery Lake Fletcher Lake Lake Major Pockwock Lake

Method Validation: Various Treatment Plants in Nova Scotia - peCOD and DOC

y = 3.04x R² = 0.91

0.0

5.0

10.0

15.0

20.0

25.0

0.0 1.0 2.0 3.0 4.0 5.0 6.0 7.0

peC

OD

– m

g/L

DOC – mg/L



Method Validation: Various Treatment Plants in Nova Scotia – peCOD and SUVA


y = 4.27x - 2.26 R² = 0.84

0.0

5.0

10.0

15.0

20.0

25.0

0.0 1.0 2.0 3.0 4.0 5.0 6.0

peC

OD

– m

g/L

SUVA – mg/L/cm3


Case Study: Biofiltration Monitoring

Biofiltration Monitoring : Background

•  Direct filtration drinking water treatment plant underwent conversion to biofiltration through removal of pre-chlorination

•  Conversion resulted in –  Reduction in HAAs (~40-60%) and THMs (~20-60%) –  Increase in bioactivity on the filter media

•  40 ng ATP/cm3 to 200-300 ng ATP/cm3

•  However, limited DOC removal across the filter occurred, making it difficult to assess treatment performance

Decrease in THM and HAA concentrations as a result of conversion

Figure adapted from: Stoddart, A. K., & Gagnon, G. A. (2015). JAWWA.

0

20

40

60

80

26-Feb-11 14-Sep-11 1-Apr-12 18-Oct-12 6-May-13 22-Nov-13 10-Jun-14 27-Dec-14

THM

—µg

/L

Filtration Biofiltration

0 10 20 30 40 50 60


HA

A—

µg/L


Biofiltration Monitoring : Approach

•  Monitored NOM surrogates (TOC, DOC and peCOD) at 3 locations for a period of 9 months

Figure adapted from: Stoddart, A. K., & Gagnon, G. A. (2015). JAWWA.

Sample Points

1 2 3

Effect of Flocculation

•  Limited removal of TOC –  TOC: 5 ± 4% –  Includes flocculated material

•  Similar removal of DOC and peCOD –  DOC: 31 ± 4%

•  Does not measure flocculated material (0.45 µm filtration as sample preparation)

–  peCOD: 32 ± 3% •  Assumed to measure only

soluble portion 0%

5%

10%

15%

20%

25%

30%

35%

40%

TOC DOC peCOD

Perc

ent R

emov

al

Flocculation Raw Water to Biofilter Influent

Error bars represent 95% CI

Effect of Biofiltration

•  Greatest average removal of TOC –  TOC: 29 ± 4% –  Flocculated material

filtered out •  Limited average

removal of DOC –  DOC: 2 ± 1%

•  More peCOD removal –  peCOD: 19 ± 5%

0%

5%

10%

15%

20%

25%

30%

35%

TOC DOC peCOD

Perc

ent R

emov

al

Biofiltration Biofilter Influent to Biofilter

Effluent


Effect of Flocculation and Biofiltration

NOM Surrogate Raw Water

Flocculated Water

Biofiltered Water

TOC—mg/L 3.16 ± 0.13 3.00 ± 0.16 2.06 ± 0.07

DOC—mg/L 3.04 ± 0.34 2.07 ± 0.06 2.09 ± 0.12

peCOD—mg/L 8.51 ± 0.55 5.90 ± 0.46 4.64 ± 0.42

Effect of Flocculation and Biofiltration

NOM Surrogate Raw Water

Flocculated Water Removal Biofiltered

Water Removal

TOC—mg/L 3.16 ± 0.13 3.00 ± 0.16 0.16 2.06 ± 0.07 0.94

DOC—mg/L 3.04 ± 0.34 2.07 ± 0.06 0.97 2.09 ± 0.12 -‐0.05

peCOD—mg/L 8.51 ± 0.55 5.90 ± 0.46 2.61 4.64 ± 0.42 1.26

Treatment Train: Combined Effect of Flocculation and Biofiltration

0%

10%

20%

30%

40%

50%

60%

TOC DOC peCOD

Perc

ent R

emov

al

Flocculation Biofiltration



SUVA Expected DOC Removal Using

Alum

>4 >50%

2-4 25-50%

<2 <25%

Source water SUVA:

3.4 ± 0.1

Expected DOC removal with alum1:

25-50%

Table. Adapted from Edzwald and Tobiason, 1999; 1Edzwald and Tobiason, 1999


0%

10%

20%

30%

40%

50%

60%

TOC DOC peCOD

Perc

ent R

emov

al



Physical/chemical Removal: ~35%


0%

10%

20%

30%

40%

50%

60%

TOC DOC peCOD

Perc

ent R

emov

al



Physical/chemical Removal: ~35%

Bio-oxidative Removal (?):

~15%

Decrease in THM and HAA concentrations as a result of conversion

Figure: Stoddart, A. K., & Gagnon, G. A. (2015). JAWWA.

0

20

40

60

80


THM

—µg

/L


0 10 20 30 40 50 60


HA

A—

µg/L


Does removal of these compounds translate to improved DBP control?

Conclusions

•  peCOD can measure NOM rapidly, at low concentrations and without the use of hazardous chemicals

•  peCOD is an appropriate bulk NOM parameter

•  The use of peCOD to monitor biofiltration may provide additional information on NOM removal and subsequent biofilter performance to compliment other NOM surrogates

Documents

Photoelectrochemical Chemical Oxygen Demand …®-COD...Civil and Resource Engineering Dalhousie University Photoelectrochemical Chemical Oxygen Demand Analysis in Drinking Water AminaStoddart