fecal sludge to energy in a prototype supercritical water ...fecal sludge to energy in a prototype...

Fecal Sludge To Energy in a Prototype

Supercritical Water Oxidation Reactor

Marc A. DESHUSSESDepartment of Civil and Environmental Engineering

Duke University, Durham, North Carolina

Open defecation… = everyday business

for about 1,100,000,000 people

2.4 billion people do not have access to

“improved” sanitation

© Bill & Melinda Gates Foundation | 4

Source: WSP analysis, using BMGF funded research

Illegally

dumped

Not effectively

treated

Effectively

treated

Drainage systems Receiving waters

9% 9% 9%1%

Leakage

Unsafely emptied

Safely emptied

Left to overflow

or abandoned

69%1%

Residential environment

79%On-site

facility

20%WC to

sewer

1%Open defecation

Untreated sludge ends in the environment (Dhaka, Bangladesh)

“INSTITUTIONAL” OPEN DEFECATION

2%

98%of fecal sludge

unsafely disposed

2%of fecal sludge

safely disposed

WSH STRATEGY

AND PRIORITIES

Confined Animal Feeding Operations (CAFOs)

= as much waste as a small town

Issues: organics and nitrogen

management, methane and odor

emissions, antibiotic resistant

bacteria, very low cost margins

• 8600 hogs

• 1400 m3 manure/week

• 17 tons COD/week

• 1.6 tons N/week

• 0.6 tons P/week8600 m3

anaerobic digester

Feces: 70-520 g/(p day) ~ 80% moisture

• Fats (5-25%)

• Carbohydrates (10-30 %)

• Nitrogenous materials (2-3%)

• Minerals (5-8%)

• Bacteria and bacterial debris (10-30%)

Where all pathogens and most of the energy is

~80 gdry , 107 g COD, ~2 g N, 1.6 MJ / (p day)

Content of Fecal Waste

Urine: 0.6 – 1.1 L/(p day)

•Organic salts (38%)

•Urea (36%)

•Organic compounds (13%)

•Ammonium salts (13%)

Contains most of the nitrogen ~7 gN/(p day)

and P ~1gP/(p d)

~440 W h/(p d)

This is a 87 kWh

dump!!!

Our Vision: Omni Processor for Fecal Waste

Sanitation for the urban poor using supercritical water

oxidation (SCWO)

Pit Latrine Collection & Transport Businesses

Supercritical treatment facility, 20 ft. container for ~1000 people

(~450-530 kWh/d)

In supercritical water, organics are rapidly

oxidized (in seconds) resulting in heat, and CO2

Benefits

o Very fast reaction (sec.)

o High conversion

to CO2 + clean water

o No SOx, NOx or odor

o No need to dry waste

o Can co-treat haz. wastes

Technical risks

o Corrosion

o Salts deposition/plugging

Pilot unit at Duke

- Heat and energy recovery

- Metallurgy and corrosion

- Process control

- Slurry pumping

System Characteristics

Basic characteristics

• 100-150 kg dry/day

• 1-2 m3/day

• Assume feed ~7-15% solids

• Reactor ID: 19 mm

• Reactor length: 4.0 m

• Heat exchanger length: 39 m

• Reynolds #: 30,000 – 60,000

• Residence time in reaction

section = 2.5 to 4.5 seconds!

Anti corrosion and plugging

measures

• High Re number, slight down

slope

• Minimize transition zones

• Reactor and part of heat

exchanger in Inconel 625

• Tandem HEPS

• Periodic maintenance

Other

• Startup with isopropyl alcohol (IPA)

but any other liquid fuel will work

• Air is used as oxidant

Pilot unit construction

Process Flow Diagram

Fecal sludge

(+co-fuel)

Air

(Furnace)

ReactorGas-liquid-solid separators

CO2, N2, O2

Clean water out

Water loop

Solids

Air-water mixture

0

10

20

30

40

Heat to 400 C

supercritical

Heat of

combusion

Po

wer

(k

W)

50 kgwet/h, 10% solids basis

Air-water preheat

Pilot unit construction

Fecal

Simulant

(lab only)

High solids-content (16%)

secondary sludge

Ash content: 24%

HHV: 15 MJ/kgdry

SCWO Feedstocks Processed

Isopropanol

(IPA)

Starter and

model fuel Dried chicken manure

HHV: 12 MJ/kgdry

Used as surrogates for

human fecal waste

(8-15% solids)Dog feces

HHV: 15.7 MJ/kgdry

Test run with 1.3% isopropanol

Typical IPA run: 99.9% removal

System Characterization with IPA

Effect of n: Stoichiometric factor of O2

84

86

88

90

92

94

96

98

100

102

0.0 0.5 1.0 1.5 2.0 2.5

Re

mo

val (

%)

O2 Stoichiometric factor (-)

COD Removal

IPA Removal

System Characterization with IPA

Calorific

content of

feed

increases

Secondary Sludge Treatment

Feed

EffluentEffluent

After settlingFeed

Biosolids as

received

(14 MJ/kg dry)

Slurry feed:

4.3% biosolids

9% IPA

Secondary Sludge Treatment

• IPA run - 3.2

MJ/kg

• Sludge run –

2.4 MJ/kg from

IPA + 0.8 MJ/kg

from sludge

T = 586 ± 4.9 °CT = 577 ± 12.2 °CT = 475 ± 17.5 °CT = 462 ± 7.3 °CT = 219 ± 0.8 °CT = 204 ± 8.0 °C

T after reaction

Secondary Sludge Treatment Summary

Removal (%)

99.97

98.12

98.60

Influent Effluent

Analysis (9.6% IPA) Steady State

COD (mg/L) 183667 50

IPA

Removal (%)

99.97

Influent Effluent

Analysis (3% sludge + 9% IPA) Steady State

COD (mg/L) 214000 70

Total N (mg/L) 10875 200

NH3 (mg/L) 443 17.6

NO3-(mg/L) 183 15.9

NO2-(mg/L) 14.9 0.4

PO4-3

(mg/L) 4930 67.9

pH 6.8 7.02

Conductivity (?S/cm) 2560 659

Sludge +IPA

Dog Feces Treatment Slurry feed pre-processing:

14% solids (15.7 MJ/kg dry)

Effluent

Feed10% solids

+ 4% IPA

Fresh Dog Feces Treatment Summary

Removal (%)

99.97-99.85

95.32-91.70

99.91-99.56

Influent Effluent

Analysis (10% sludge + 4% IPA) Steady State

COD (mg/L) 192276 65-280

Total N (mg/L) 4704 220-420

NH3 (mg/L) 627.2 185-325

NO3-(mg/L) 98 0.3-0.8

NO2-(mg/L) 22.54 0.04-0.54

PO4-3

(mg/L) 14543.2 13.4-63.9

pH 5.95 -

Conductivity (µS/cm) 4500 -

Dog poop +IPA

Treatment of Micro PollutantsExperimental Approach

• Spiked trace contaminants during a run

• Used high concentrations of Triclosan, Acetaminophen, and

Ibuprofen

• Run with spiked IPA first, then spiked dog feces and IPA

Results

• Ibuprofen and acetaminophen (10 mg/L each) – not analyzed yet

• Triclosan:

Relevant concentration: 0.5 – 1 µg/L

Concentration spiked: 100 µg/L

Concentration in effluent (IPA treatment): ND at < 0.1 µg/L

Concentration in effluent (dog feces + IPA treatment ): < 0.1 µg/L

Removal > 99.99%

Modeling Studies• Develop a conceptual model of the process (with reaction

kinetics, mass and energy balances)

• Use model to:

– Conduct data analysis

– Explore process sensitivity

– Process optimization and design

• Methods: Aspen Plus

– Peng-Robinson equation of state

– Plug flow kinetic reactor (Rplug)

– Heat exchangers sizing using the Exchanger Design and

Rating (EDR) tool

2𝐶3𝐻8𝑂 + 9𝑂2 → 8𝐻2𝑂 + 6𝐶𝑂2𝐶3𝐻8𝑂 + 3𝑂2 → 4𝐻2𝑂 + 3𝐶𝑂2𝐶𝑂 + 𝑂2 → 2𝐶𝑂2

𝑟𝑖= 𝑘𝑒 Τ−𝐸𝑎 𝑅𝑇𝐶𝐼𝑃𝐴𝐶𝑂2

Reactions (IPA)

Modeling the Heat Exchangers

Exp. Model

U (W/m2-K) 196-205 367

Q (kW) 47-49 51

LMTD (K) 114 65

Shell side

Tube side

Exp. Model

U (W/m2-K) 90-124 89

Q (kw) 3.6-4.9 4.1

LMTD 33 39

Modeling the Heat Exchangers

Shell side

Tube side

Adding Heat Losses to the System

Model vs. Experiments

We are currently using the model

to optimize process design and

operation using a factorial

approach

Motor

Size (kW)

Duty

Cycle

Actual

Power (kW)

Motor

Size (kW)

Duty

Cycle

Actual

Power (kW)

Furnace 15 5% 0.75 Furnace 15 90% 13.5

Air Compressor 11.2 80% 9.0 Air Compressor 11.2 80% 9.0

Water Pump 3.7 50% 1.9 Water Pump 3.7 50% 1.9

Biomass Pump 11 15% 1.7 Biomass Pump 11 15% 1.7

PC Pump 0.56 60% 0.34 PC Pump 0.56 60% 0.34

PLC and Sensors 1 100% 1 PLC and Sensors 1 100% 1

Steam Generation -4 100% -4 Net Electric Power Draw (kW): 27.3

Gaseous Expansion -0.1 100% -0.1 Gasoline Equilivance* (gal/day): 49

Net Electric Power Draw (kW): 10.4

Gasoline Equilivance* (gal/day): 18

Current and Projected Energy Balance

Currently not

autothermal:

But we have 17

kW heat loss in

final cooling and

~10 kW heat

losses around

the reactor

Optimized design:

• Turn of furnace (autothermal)

• Recover energy from gas expansion (3-6 kW)

Expected draw ~6-10 kW

Produce 5-10 kW as heat

Some Lessons LearnedDesign

• Pumping slurries at low flow and high pressure is a challenge

• Compression fittings in temperature zones

above 350 °C = high failure rate

Operation

• Capillary depressurization system is difficult to control during

start up and shut down but

work well otherwise

• Foam in the HEPS can affect

level sensor reading

Modeling

• Losses play a crucial role,

requires some tricks in Aspen

Concluding Remarks: Why I am Optimistic…

• SCWO achieves both waste treatment and pathogen

control extremely fast. We can even co-treat

hazardous wastes and produce clean water, without

odor, SOx, or NOx emissions…

• Selling “high value added” by products can be a driver

Sell a 10 L shower 5-10 cents

= $1.7-3.4 per kWh!

• But many challenges remain…

AcknowledgmentsGates Foundation and USDA

for fundingDeshusses lab

marc.deshusses@duke.edu http://sanitation.pratt.duke.edu/