October 30, 2012

Joe Tesar, Quantalux, LLC Dana Kirk, MSU Department of Biosystems and Agricultural Engineering John Willard, Quantalux, LLC

Anaerobic Digester System Failure Analysis, Corrective Actions and BMPs

Quantalux and ADREC/MSU

Quantalux LLC: Developers of innovative renewable energy technologies

Current Focus: - Small-scale Biogas Technology Development

- Thermal/Energy Modeling of AD Systems

- Value-enhancement for Feedstock Selection

Anaerobic Digester Research and Education Center (ADREC) Based at Michigan State University, ADREC has a fully-equipped

teaching & research facility specifically tasked to enhance digester

technology.

• Feedstock, Additive and Biogas Characterization

• Pilot Digesters and Test equipment

• Experienced Staff and Students

AD System Failure Analysis Goal: Use past history to improve system performance

• Approach:

– Use documented data to identify Cause and Frequency of Digester Failures

– Seek trends and ID common problems

• Sources include published info and personal interviews

– Case Study Data from Dairy Farms across the US

– Sensor Data from AD systems in NY State

By “Failure”, we mean: Failure to Perform to Spec, Failure to meet Cash Flow, etc.

Methodology and Resources

• 25 Dairy Farms in the US

− Published Case Studies

− Phone interviews

• NYSERDA CHP Operational Data (4 farms)

• Failures divided into 5 categories

• Document solutions and corrections that took place

Agricultural Digesters in the US

• US EPA AgSTAR – 192 operating anaerobic digesters

on farms in the U.S.

• Dairy farms are leading adopters – 159 systems (82% of farm projects)

Majority of Digesters are Plug-Flow and Complete Mix. Data and graphic from: http://www.epa.gov/agstar/projects/index.html

Fixed film, 2%

Induced blanket reactor,

2%

Other, 5%

Covered lagoon,

9%

Horizontal plug flow,

13%

Complete mixed, 32%

Mixed plug flow,

37%

All On-farm Digester Systems in the US

Fixed film, 1 Induced blanket

reactor, 1 Mixed plug flow, 1

Covered lagoon, 2

Other, 2

Complete mix, 9

Horizontal plug flow, 9

Our Sample Set for the Study

This sample set includes published case studies

*U.S. EPA AgSTAR

Study Statistics

Sample Size 25 Farms

Percentage of Operating Agriculture Digesters*

13%

Farm Size Range (Dairy Cows) 100 to 7500

Total Capacity ~ 11 MW

Total Dairy Cow Count 46,000+

Cogeneration 84% (21/25)

Thermal Energy Only 16% (4/25)

Farms Included in Failure Analysis

In order to keep the focus is on the resulting data (and not on the specific farm operator), we are not identifying farms by name. Instead, we use Farm 1, 2, 3, etc.

Failure Categories (5x)

Site Planning & Design – includes site plan development and integration into existing facilities

Engineering – includes all engineering related activities (civil, structural, electrical and mechanical)

Construction & Equipment – includes construction quality and equipment selection for the digester

Biogas Utilization – includes equipment selection and system integration of the biogas utilization system

System Control & Operation - monitoring and control (including personnel issues)

Failure Breakdown in our Sample Set

0%

10%

20%

30%

40%

50%

60%

70%

80%

90%

100%

Appx. 60% of failures arise in design and construction phase.

Appx. 40% of failures occur after the system is completed.

System Control & Operation

Biogas Utilization

Construction &

Equipment

Engineering

Site Planning & Design 16%

17%

25%

16%

25%

Failure examples

Site Planning & Design – includes site plan development and integration into existing facilities

Issue Corrective Action

Excess heat loss to soil and groundwater below the pad of the digester vessel.

Provide plenty of insulation to avoid heat loss. Require calculations for expected heat transfer.

High solids content of the manure required additional dilution water to allow the material to be transferred from the barn to the digester.

Analyze manure characteristics to assess flowability (solids content & viscosity)of the manure. Operator must follow engineering recommendations for transferring manure with a wide range of solids characteristics.

Failure examples

Engineering – includes all engineering related activities (civil, structural, electrical and mechanical)

Issue Corrective Action

Gas leakage from the concrete hard top of the digester resulted in raw biogas emissions causing odor problems plus loss of energy.

Post-construction validation and verification (V&V) needed – initial pressure tests plus leak monitoring during system ramp-up.

: Difficulty in heating the manure entering the digester caused poor biogas production. Frozen manure clogged pipes.

Mild preheating (biogas-fired) using a pipe-grid melted frozen chunks of incoming manure and maintained material flow.

Failure examples

Construction & Equipment – includes construction quality and equipment selection for the digester

Issue Corrective Action

CHP units failed after 3000 hr Input to CHP requires enhanced scrubbing (H2S removal) and a water separator. Periodic testing of H2S levels into CHP essential.

Mechanical issues with Genset exceeding benefit from electrical production. Repairs included valves, muffler and biogas lines.

Inadequate biogas scrubbing was corrected and regular maintenance procedures for Genset were instituted.

Failure examples

Biogas Utilization – incl. equipment selection & system integration of the biogas system

Issue Corrective Action

Boiler Corrosion due to excess H2S levels

Provide for continuous H2S monitoring AND train personnel on scrubber/filter operation and maintenance.

Biogas pipe connecting the flare was undersized (3”) and inadequate when the flare was the only outlet for the gas.

Biogas piping must be sized for anticipated flow PLUS an appropriate safety factor.

Failure examples

System Control & Operation - monitoring and control (including personnel issues)

Issue Corrective Action

Variable biogas pressure and methane concentration led to poor performance by microturbines.

Designers must adhere to technical specifications for all equipment.

Periodic operation of mixers resulted in crust formation (and decrease in biogas)

Optimize the duty cycle for mixers for a given feedstock to assure best mixing while using minimum energy.

Difficulty maintaining digester temperature

Prewarm frozen manure, and also assure that all temperature sensors are calibrated and functioning properly.

Data from Operational Anaerobic Digesters: Trends and Observations

• Data from 4 farms in NY State:

– Sensor Data: Generator Down-Time, Gas Flows, Facility Energy, Useful Heat Recovery (CHP) and Ambient Temperatures

– On-site performance reports (quarterly updates)

• From the data, we calculated:

– Average Generator Output (kWh)

– % Biogas in Flare

– Contract Capacity

– Revenue Lost due to Down-Time (assuming $0.10/kWh)

1) Dry manure causing piping problems = low biogas production

2) Unknown event

3) Unknown event

4) Genset down for cleaning. Repairs to digester heating and system recirculation pumps

$14

$12

$10

$8

$6

$4

$2

$0

0

100

200

300

400

500

600

700

1 4 7 10 13 16 19 22 25 28 31 34

Lost

Re

ven

ue

/mo

($

1,0

00

)

Generator Hours Down

Average Generator Output

Contract Capacity

Capacity

Revenue Lost

Ho

urs

/ k

W A

vg

Farm 1: Plug Flow – 1800 Cow – 500 kW

1 2 3 4

Documentation of generator downtime is necessary to determine cause of failure !

Lost revenue from Generator Hours Down accumulates over term of the project

$18

$16

$14

$12

$10

$8

$6

$4

$2

$0

0

100

200

300

400

500

600

700

5 8 11 14 17 20 23 26 29 32 35 38 41 44 47 50 53 56 59 62 65 68 71 74 77 80 83

Lost

Rev

enu

e/m

o (

$1

,00

0)

Generator 1 Downtime

Generator 2 Downtime

Contracted Capacity

Combined GeneratorOutputCombined Capacity

Ho

urs

/ k

W A

vg

Farm 2: Complete Mix – 360 Cows – 250 + 225 kW

1) Second generator added

2) The oil cooler on the existing genset went bad and contaminated the antifreeze with oil.

3) Overheating issues associated with the inability to reject enough heat in summer months.

4) Existing engine rebuild, expected down time was 30 day, actual = 3 months

5) Existing engine required new radiator

6) Change in feedstock results in lower biogas production

1

2 3 4 5

6

After four years of consistent operation, significant problems with the original Genset!

Revenue Lost

$20

$18

$16

$14

$12

$10

$8

$6

$4

$2

$0

0

100

200

300

400

500

600

700

1 4 7 10 13 16 19 22 25 28 31

Lost

Re

ven

ue

/mo

($

1,0

00

)

Ho

urs

/ k

W A

vg

Months

Generator Hours Down

Generator Output Average

Generator Capacity

Contracted Capacity

Revenue Lost

Farm 3: Plug Flow – 1100 Cow – 416/500 kW

1) Initial issues getting the generator up and running.

2) Generator re-tuned to run at 500kW

3) Oil Change and header replacement (5 days of downtime)

1

2

3

Meeting design expectations is important for considering contract capacity.

Operational Month

$16

$14

$12

$10

$8

$6

$4

$2

$0

0

100

200

300

400

500

600

700

1 4 7 10 13 16 19 22 25 28 31 34

Lost

Rev

enu

e/m

o (

$1

,00

0)

Generator Hours Down

Capacity

Average Generator Output

Revenue Lost

Ho

urs

/ k

Wh

Avg

Farm 4: Plug Flow - 2700 Cows – 500 kW

1 2 3

1) Manure shortage

2) Headers changed

3) UNK Event

Operational Month

Operational problems within the first 3 years led to significant loss of revenue for the AD owner.

20

0

100

200

300

400

500

600

700

800

1 5 9

13

17

21

25

29

33

37

41

45

49

53

57

61

65

69

73

77

81

85

89

93

97

10

1

10

5

10

9

11

3

11

7

12

1

Ho

urs

Operational Month

GenSet Downtime – All A.D.s in NY State

Complete Mix

Complete Mix

Complete Mix

Complete Mix

Mixed

Mixed

Plug Flow

Plug Flow

Plug Flow

Plug Flow

Plug Flow

Plug Flow

Plug Flow

Plug Flow

Plug Flow

Plug Flow

MONTH

1 WEEK

2 WEEKS

Overhaul

Minor Maintenance

Major Maintenance

Note that the data on GenSet down-time is very well correlated with key maintenance events

Key Take-Aways

21

Different Problems : Short Term vs Long Term • Over initial 2-5 years

Engineering /Design issues arise, and must be addressed. Avoid problems with good contract/contractor.

• Over 20+ yr design lifetime – Different issues Genset maintenance/repair

• Typically takes longer than expected • Result = Additional lost revenue ( -$$)

Feedstock management critical (no foam, no additives)

Final Thoughts

22

• Data and case studies can be valuable to identify typical problems Sensor data is highly detailed

• BUT incomplete without accompanying report. • All failure events must be documented to ID “cause”

Recommend: • The industry develop a shared database of information

on digester performance. Broad view of digester systems across the US

More data on performance => better reliability and improved finances.

Best management practices (BMPs)

23

KEY Long term strategy Assure Continuous Gen-Set Operation • Train Personnel

Define and Follow to Maintenance Schedules Monitor feedstocks and biogas production Follow pre-defined Standard Operating Procedures

• Pre-stage spare parts for rapid replacement Once identified, repairs must be accomplished and biogas

production returned to normal

• Pre-plan major overhauls Predictable events Large potential for lost revenue

Acknowledgements

24

We gratefully acknowledge support from the US Department of Agriculture, National Institute for Agriculture (NIFA) for this analysis under SBIR Contract # 2011-33610-30777 (Phase II SBIR to Quantalux, LLC “A Biogas Heat Engine for Small to Mid-sized Farms”)

Contact: Dr. Charles Cleland, ccleand@nifa.usda.gov