Re-engineering Carbon Cycling for Resource Recovery from

Waste Streams

Kartik Chandran

Columbia University

Bioeconomy 2017

Washington, D.C.

July 12th, 2017

This presentation focuses on

Options for carbon recovery to fuels and chemicals

(think beyond CH4)

Resource efficient options for wastewater

treatment and sanitation

AMMONIA

OXIDIZING

BACTERIA

Ammonia

Nitrite

Methane

Methanol

O2

Water

Oxidation of ammonia as the

primary energy source for energy

metabolism

Oxidation of methane via co-

metabolism, without net energy

synthesis

Foundation for resource recovery through anaerobic

C-conversions

Complex organic polymers

Sugars, amino acids

VFA

Acetic acid

Hydrolysis

Acidogenesis

Acetogenesis

Methanogenesis

Anaerobic

Digestion

HRT > 10 dMethane

Complex organic polymers

Sugars, amino acids

VFA

Acetic acid

Hydrolysis

Acidogenesis

Acetogenesis

Anaerobic

Fermentation

HRT ~ 2 d

• Fermentation (still using mixed microbial consortia) is more advantageous than just anaerobic digestion

• Fermentation can be incorporated into existing digestion processes

Fermentation as a platform for resource recovery

I. Strategies to increase carboxylic acid production

from anaerobic fermentation of food waste

Variables tested (T=35oC)

• Organic loading rate

•

• pH

•

10 and 25 kg COD/m3-day

6.5, 7.0, 7.5, 8.0, 9.0

• Solids retention time

• 2 d, 4 d, 6 d

Performance indicators

• sCOD yield

• Total carboxylic acid yield

• VFA yield

• TCA concentration

• VFA concentration

Yield = Effluent (sol) – Influent (sol)pCODin

Pavlakis et al., unpublished

No difference in yield for OLR 10 vs OLR 25

…but VFA concentrations are higher in OLR25

Fermentation at higher pH did not improve

yield or effluent concentrations

Increasing fermenter SRT at alkaline pH

increased effluent VFA concentrations

and fractions

Pavlakis et al., unpublished

Non-TCA sCOD needs to be

characterized to expand valorization

options

Elevated pH and SRT increased selectivity for

acetate production

100908070605040302010

0HRT 2dpH=6.5

HRT 2dpH=7

HRT 2dpH=7.5

HRT 2dpH=8

HRT 2dpH=9

HRT 4dpH=9

HRT 6dpH=9

Pe

rcen

to

fV

FAp

rod

uce

d

Fermenter pH

Acetic (AA)

Valeric (VA)

Propionic (PA)

Succinic (SA)

Formic (FA)

Other LC Acids

Butyric (BA)

Pavlakis et al., unpublished

Annavajhala et al., unpublished

Impact of configuration on function and metabolismComplement of pathways present in anaerobic fermentation processes

Challenges and opportunities

• Anaerobic fermentation is a fairly flexible process (feed

composition, operating strategies)

– Optimization possible through appropriate process engineering

• Feedstock pre-processing

– Homogenization, pre-hydrolysis

• VFA separation

– Low tech – elutriation (washing)

– Membrane separation, pervaporation and more

• VFA speciation and concentrations for downstream conversion

– 100% purity not entirely essential

Organic feedstock

Anaerobic fermentation to produce volatile fatty acids (VFA)

Convert VFA to lipids

Harvest and extract

lipids Convert lipids to …

II. Organic feedstocks to lipids and more

0

10

20

30

40

50

60

70

C 14:0 C 16:0 C 16:1 C 18:0 C 18:1 C 18:2 C 18:3 C 20:1

Rela

tive

% o

f F

AM

E26 mgN/L

52 mgN/L

130 mgN/L

260 mgN/L

1300 mgN/L

Fermenter VFA

Glucose

Major fatty acids accumulated are palmitic (C16:0), oleic (C18:1), and linoleic acid

(C18:2)

Most similar to palm oil

Composition of lipids accumulated by Cryptococcus albidus

Vajpeyi and Chandran, 2015

What else can C. albidus accumulate (or do)?

Under what conditions?

13Total size ~ 25 Mbp, Genbank accession number LKPZ 00000000.1 Vajpeyi and Chandran, 2016

Challenges and opportunities

• Potential to enhance product yield and rate

– Process engineering

– Genetic manipulations

– Impact of co-culturing bacteria and yeast

• Cell separation and extraction

– Re-utilization of chemicals

• Re-utilization of ‘waste’ biomass (?)

– Digestion

– Co-processing with other biomass

– Nutrient extraction

III. A. Internal use of VFA for enhanced BNRDual-Phase Digestion and Fermentation of AS

PDS fermentation and storage at 26th Ward WPCP in

New York City, 2002

• Fermentation of PDS to

produce VFA

– Used mainly for denitrification

– Kinetics higher than MeOH

III. B. Faecal Sludge to Biodiesel in Kumasi, GH

• Concomitant oxidation of CH4 and CO2 fixation

• Prospect of combining C &N cycles

AMMONIA

OXIDIZING

BACTERIA

Ammonia

Nitrite

Methane

Methanol

O2

Water

Oxidation of ammonia as the

primary energy source for energy

metabolism

Oxidation of methane via co-

metabolism, without net energy

synthesis

Sewage sludge to methanol

Taher and Chandran, ES&T, 2013

WE&RF Paul Busch Award, 2010

CO2

Biomass

III. C.

Biological

conversion

Concluding remarks

Wide variety of endpoints (chemicals,

fuels..) possibleDisrupting or complementing conventional

pathways to biofuels

Links to other applications needed

and possibleResource efficient options for wastewater

treatment and sanitation

AMMONIA

OXIDIZING

BACTERIA

Ammonia

Nitrite

Methane

Methanol

O2

Water

Oxidation of ammonia as the

primary energy source for energy

metabolism

Oxidation of methane via co-

metabolism, without net energy

synthesis

Channeling anaerobic C-conversions through

SC-FA offers attractive flexible prospects for

resource recoveryDetailed understanding in conjunction with reductionist

approaches needed to advance implementation

Discussion

Shashwat Vajpeyi, Medini Annavajhala, Justin Shih, Ato

Fanyin Martin, Edris Taher, Yu-Chen Su, Huijie Lu,

Vladimir Baytshtok, Jorge Santodomingo, Vikram Kapoor

Kartik Chandran

Professor

Director, Wastewater and Climate Change Program

Director, Columbia University Biomolecular Environmental

Sciences

Email: kc2288@columbia.edu

Phone: (212) 854 9027

URL: www.columbia.edu/~kc2288/

Significance

• As we attempt to move away from ‘digestion’ or ‘fermentation’ towards carbon

recovery and biorefining, we need more structured information

• It becomes beneficial to link

– Process configurations and operating conditions with microbial ecology, metabolic function

• and in turn with product yield and speciation

– Feedstocks (can import or mine added ones) with products

• Next steps

– Which pathways are active? mRNA, protein expression

– At what rate? 12C or 13C metabolite tracking

• Not needed for every case

– Reductionist approaches needed to enhance translation

Novel and flexible platform to convert

a variety of organic ‘waste’ streams to

biodiesel or other lipid based

commodity chemicals

Not reliant upon inherent lipid

content- other organic classes can be

converted to lipids

Fermentation to produce

VFA

Municipal solid waste

Faecal sludge

Domestic and Food waste

Agricultural waste

Animal by-product waste

Biodiesel and other commodity chemicals

• For biodiesel as the preferred end point, reliance upon agricultural outputs is reduced or eliminated

• Links clean water production with energy and chemical recovery

• Process enhancements through fundamental understanding of the biological platform

• Cost?

Significance

0

500

1000

1500

2000

Formate Acetate Propionate Butyrate Succinate Valerate

VFA

co

nce

ntr

atio

n

(mgC

OD

/L)

2 Day HRTStep-feed

Non-step-feed

• Largely similar VFA

speciation and yields

observed across HRTs

• Acetate, propionate and

butyrate were the

dominant VFA

0

500

1000

1500

2000

Formate Acetate Propionate Butyrate Succinate Valerate

VFA

co

nce

ntr

atio

n

(mgC

OD

/L)

4 Day HRTStep-feed

Non-step-feed

0

500

1000

1500

2000

Formate Acetate Propionate Butyrate Succinate Valerate

VFA

co

nce

ntr

atio

n

(mgC

OD

/L)

8 Day HRTStep-feed

Non-step-feed

Insights from fermentation experiments

• Increasing OLR can increase TCA concentrations –

potentially limited by pCOD hydrolysis

• Increasing pH and SRT can boost acidogenesis

– Tradeoff could be increased CH4 production

• Maximizing hydrolysis and acidogensis requires an effective

strategy to separate products from reactants

Impact of configuration on community structure

Annavajhala et al., unpublished

Comparison of microbial community profiles in SBR and SFR at 8d, 4d, and 2d HRT; genera included have been assigned >= 50 reads per million (RPM) for

at least one HRT; dark blue or green color indicates assignment of >= 1,000 RPM; light blue or green color indicates assignment of >= 50,000 RPM; dot size

scaled to RPM values; shading reflects phylum-level relationships

Acidogenesis improved at elevated pH and SRT

0

2,000

4,000

6,000

8,000

10,000

12,000

2 day

pH 7.0

OLR 25

2 day

pH 7.5

OLR 25

2 day

pH 8.0

OLR 25

2 day

pH 9.0

OLR 25

4 day

pH 9.0

OLR 25

6 day

pH 9.0

OLR 25

Co

nce

ntr

atio

n

(mg

CO

D/L

)

sCOD (not TCA) TCA (not VFA) VFA Concentration

Non-TCA sCOD needs to be characterized to expand valorization options

Pavlakis et al., unpublished