1

Development of a Calculator for the Techno-economic Assessment of Anaerobic Digestion Systems

Anthony Lau, Sue Baldwin and Max WangDepartment of Chemical & Biological EngineeringUniversity of British Columbia

19th Annual National Composting ConferenceOctober 1, 2009

1

Introduction

Objectives

Methodology

Results and Discussion

Conclusions

Acknowledgement

2

Introduction

3

Stages of Anaerobic Digestion

IBBK, International Biogas Study Tour 2006 Southern Germany and Austria (Cited by Electrigaz Technologies Inc. 2007 “Feasibility Study - Anaerobic Digester and Gas Processing Facility in the Fraser Valley, British Columbia”

Biogas generation potential of substrates

Biogas generation potential,

m3/tonne substrate (wet mass basis)

Substrate Electrigaz Technologies, 2007

Preusser, 2006

Kramer and Krom 2008

cow manure 25 (9% TS) 25 25

pig manure 25 (7% TS) 35 30

potato/vegetable

waste

60 (10% TS) 70 39

corn/grass silage 175 (25% TS) 200 185

food waste 225 (20% TS) 175 265

fats and grease 500 (50% TS) 980 961

Substrate Biogas Yield Units Reference cattle slurry pig slurry poultry slurry whey food waste grass silage

0.25-0.30

0.20-0.50 0.35-0.60 0.35-0.80 0.25-0.60 0.56

m3/kg VS Seadi, 2001

dairy cattle manure beef cattle manure pig manure mixed food waste fruit/vegetable wastes corn silage

0.25 0.55 0.53 0.80 0.30-0.80 0.65

m3/kg VS Ward et al. 2008

soup processing waste cafeteria waste kitchen waste fish farm waste grease trap

112 (21.5% TS) 150 (23.5% TS) 53 (9.7% TS) 472 (55.8% TS) 275 (29.4% TS)

m3/tonne feed (wet mass basis)

Zhang et. al. 2007

Biogas yields (lab-scale and pilot-scale AD studies)

Cattle slurry: Based on average manure generation characteristics, 0.25-0.30 m3/kg VS may be converted to 24-28 m3/tonne feed

1 HPF: horizontal plug flow; MPF: mixed plug flow; MMPF: modified MPF; CM: Complete mixed2 OFMSW: organic fraction (source separated) of municipal solid waste 3 Feedstock made up of 55-70% of cattle manure, and 6-38% of organic wastes which include some or all of

the followings: slaughterhouse waste (intestinal contents), fats, fish processing.

Substrate Facility/Location Biogas Yield, m3/tonne feed (w.b.)

Reactor type1 or AD process

Reference

Dairy manure Straus Creamery, CA Gordondale Farm, NY AA Dairy, WI Baldwin Farm, WI Sheland Farms, NY

11.0 40.0 35.7 28.0 20.0

Covered lagoon MPF HPF MMPF CM

Anon, 2004 Martin 2005 Martin 2004 USEPA, 2009 Pronto and Gooch 2008

Dairy manure/ food waste

Ridgeline Farm, NY Holsworthy, UK

73.5 40.4

CM Wet, single-step

Pronto and Gooch 2008 Beck Inc., 2004

OFMSW

Geneva, Switzerland Ameins, France Vagron, The Netherlands Wels, Austria Toronto, ON

120 150 40.8 89.5-140 95-110

Dry, single-step Wet, single-step Wet, two-steps

Beck Inc., 2004 Goldstein, 2005

Manure and organic wastes 3

Various locations in Denmark

23-98

Not specified

Gregerson et al., 1999

Power Generation (kW/cow)

1997-2002a 2002-2008b, c

PF Digester 0.08 - 0.17 0.16 - 0.21d

MPF Digester 0.15 - 0.23 0.16 - 0.28e

CM Digester (CSTR) 0.10 - 0.23 0.23 - 0.32f

Covered Lagoon n/a 0.12

Given the same technology, the greater values of power generation were generally associated with the co-digestion of manure and other organic wastes.

This table does not include data pertinent to Microgy AD systems. Their generators are over-sized with power generation ranging from 0.75-0.90 kW/cow.

a USEPA - AgSTAR Handbook b AgSTAR Program - Guide to anaerobic digesters c Cornell University – Manure management Programd mostly RCM Digesters Inc/RCM International Inc.e mostly GHD Inc.f various suppliers

Digester Configuration

At present, CSTR and MPF are the most commonly used systems.

Some Suppliers: CSTR: BTA, HAASE, RCM PF: RCM, OWS (Dranco), Alliant Energy MPF: GHD, Kompogas, BIOTHANE

CSTR

Plug Flow (PF)

MPF Digester

Mixed Plug Flow (MPF) Digester Two configurations

Rotary Two Chamber

Separation of bacteria culture

Solid Retention Time (SRT) >> Hydraulic Retention Time (HRT)

Advantages of MPF

An easy approach to separate two groups of bacteria involved

Biomass recirculated to the second tank to enhance digester performance

More efficient than CSTR and easier to operate than Plug Flow

Recall

Bacteria growth - two chamber kinetics

More About Solids Retention Time SRT vs. Hydraulic retention time HRT

Without recirculation, SRT = HRT. With recirculation, SRT is the actual

period of digestion. This is harder for CSTR and PF to

achieve due to the lack of bacteria culture separation.

Dry vs. Wet Digestion

TS > 20% Dry, TS < 15~20% Wet Wet digestion is more popular for farm-

size operations. Dry digestion may have longer HRT. Some Suppliers:

Dry: PlanET, Kompogas Wet: GHD, BIOTHANE

Mesophilic vs. Thermophilic

Around 35oC is mesophilic, over 50oC is thermophilic

Mesophilic is more popular for farm-size operations

Some Suppliers: Mesophilic: GHD, Alvesta Thermophilic: Kompogas, Microgy

Objectives and Methodology

21

Main goal of this research project:

To develop an Anaerobic Digestion Calculator that would assist farm owners in the Lower Fraser Valley of BC in making decisions on choosing suitable anaerobic digestion technologies for their own farms.

Objectives:

7. To inform potential users of the currently available technology options for both anaerobic digestion and biogas utilization; and

9. To model the selected AD and biogas utilization technology.

Desirable features of the calculator

The ability to input amounts of different types of wastes including animal, food, agricultural and municipal wastes.

A user-friendly interface for choosing from a selection of digester types and biogas utilization - cogeneration or biogas upgrading.

A robust model parameter input interface, which should provide default values for average users, but also allow advanced users to input their own parameters to match their particular feed or design.

A detailed output including all the input information, model parameters used and calculated results. Users should be able to save this output and view it as a report.

Help documentation for both basic and advanced users.

Biogas plant

Digester Effluent Storage Biogas Utilization

Calculator - Quick Interface Guide

How to Model It?

Two Schools of Approaches

Step-Wise Models

Disadvantages

A very complicated model All the parameters will have to be measured for

different types of organic materials. However, we are not interested in the concentrations

of all the materials considered in the model. Also, Excel does not have a built-in algorithm for

solving differential equations.

Monod-Like Models

What Does This Model Say?

bSK

akSS

−+

=µ

S = the concentration of substrate (mg/L)

a = the growth yield constant (mg cells or bacteria/mg substrate)

k = the maximum rate of substrate utilization (mg substrate/mg bacteria.day)

Ks = half-growth velocity (mg/L)

b = decay rate (1/day)

Focus on two types of digesters:

Completely Mixed Digester (CSTR) Modified Plug Flow Digester (MPF)

Energy Balance

Technology and Supplier Digester Co-generation Biogas Upgrading

Available Software Coefficient-based Kinetics-based

Economic Analysis

Split Biogas

Coefficient-based Works with some constant coefficients, such

as: digester volume per cow biogas production per ton of waste electrical power per cow

It provides fair estimation assuming that digestion achieves the expected conversion.

Easy to apply

Disadvantages

Coefficients vary from one source to another.

Does not distinguish between different digester types.

No prediction on “what if”, such as incomplete digestion, shorter HRT, complex waste sources.

Kinetics-based This project: Lawrence & McCarty’s Model

FarmWare 3.03.1: Chen & Hashimoto’s Model

bSKKS

S −−+

=)1(0

maxµµ

bSK

akSS

−+

=µ

Derivation

Both follow Monod-type kinetics, and are very similar if we consider:

akK

≈−1maxµ

SKKKS ≈−1

0

There is no consensus on which kinetic model is more realistic.

Both consider inhibition from initial substrate concentration and the nature of the substrate.

Lawrence & McCarty model has more industrial backgrounds and data available.

Calibrating kinetic parameters

We selected 4 sites with 5 sets of data

Haubenschild Farm: 750 cows, PF Walford College Farm: 130 cows, 160 pigs, CSTR Linsbod in Austria: 6 m3 poultry manure, CSTR Davinde Biogas Plant: 3 pig farms, 3 cattle farms,

sludge and fish wastes, 28 ton/day, CSTR

Calibration Function

Results and Discussion

42

Model demonstration/validation Predicted CH4 Reported power Predicted power generation production productiont/d x 106 kWh/yr x 106 kWh/yr

1.5

1.1

0.7

1.94

2.62

1.93

1.19Montagne Farm

1100

Blue Spruce Farm

Pleasant Valley Farm

Green Mountain Farm

680

1.30

3.20

1.80

1.40

#cows

950

1500

0.9

Modified plug flow AD systems 2006/2007. Since actual manure characteristics were unknown, default values in the calculator were used for all kinetic parameters and feed characteristics.

A Predictive Example …

A farm site with 450 dairy cows With and without off-farm wastes, such

as food waste (~ 20% by weight) Co-generation

Manure/slurry generation: 25 m3/d (or, equivalent to 0.055 m3/cow.d, ASAE Standards, 2008); DM 12.5% w.b. The manure will be diluted from DM 12.5% to 0.0%.

Digester operating temperature: 35 C (mesophilic) Average annual ambient temperature: 13.8 C Digester configuration: Diameter-to-length ratio is 1.5:5.0

For co-generation Heat recovery efficiency: 50% Power or electricity recovery efficiency: 30% Combustion or engine efficiency: 90% Utility fraction (percent co-generated power used to heat the

digester): 5%

Computed AD system performanceCSTR MPF

HRT 25 d

HRT 28 d

HRT 30 d

HRT 20 d

HRT 22 d

HRT 25 d

Digester volume, m3

1005 1126 1207 644 708 804

Biogas production CH4 production, ton/d CO2 production, ton/d

0.2 0.4

0.5 1.0

0.5 1.1

0.5 0.9

0.5 1.1

0.6 1.3

Co-generation Heat production, 106 kWh/y Power production, 106 kWh/y kW kW/cow Power purchased, 106 kWh/y

0.237 0.298 34 0.08 0.015

0.925 0.717 83 0.18 0.036

1.102 0.827 96 0.21 0.041

0.908 0.681 79 0.18 0.034

1.080 0.790 91 0.20 0.040

1.383 0.975 112 0.25 0.049

VS reduction, %

23.4 56.5 65.3 53.8 62.4 77.1

Results of economic analysesproject life 10 years; 70:30 equity/debt financing; loan interest 6% per year; MARR 10%; electricity purchased at 8 cents/kWhr and sold at 9 cents/kWh;sales revenue only due to power generated; tax rate 13.5%; CCA rate 30%

CSTR MPF HRT

28 d HRT 30 d

HRT 20 d

HRT 25 d

Co-generation Power production, kW

83

96

79

112

Revenue Electricity sale, $/yr

64485

74401

61253

87759

Costs Capital cost, $ Operating cost, $/yr Utility cost, $/yr

704436 38040 2866

784384 42357 3307

342142 18476 2722

463899 25051 3900

Cash flows Before-tax cash flow BTCF, $/yr After-tax cash flow ATCF, $/yr

23580 -5133

28737 -3235

40055 26110

58808 39899

Profitability indicators Based on BTCF Net present value NPV, $ Internal rate of return IRR, % Simple payback period PP, yr Based on ATCF Net present value NPV, $ Internal rate of return IRR, % Simple payback period PP, yr

-333553 -- 20 -510467 -- --

-356161\ -- 20 -555141 -- --

13747 11 5.5 -79175 1.7 9.5

46278 13 5.5 -84626 3.5 8.5

Capital cost estimate as a functionof number of cows

y = 7.1901x0.7737

R2 = 0.9096

y = 0.2033x1.1734

R2 = 0.8991

0

1000

2000

3000

4000

0 500 1000 1500 2000 2500 3000

Number of dairy cows

Cap

ital c

ost,

x 10

00 $

modified plug flow reactorcomplete-mix reactor

Capital cost estimate as a function of maximum power output

y = 26.917x0.7388

R2 = 0.8589

y = 7.6359x0.8753

R2 = 0.8117

0

1000

2000

3000

4000

0 100 200 300 400 500 600 700

Power generated, kW

Capi

tal c

ost,

x 10

00 $

modified plug flow reactorcomplete-mix reactor

Conclusions and Recommendations

50

The calculator software was constructed on Excel spreadsheets with simple user interfaces coded via Visual Basic applications. This makes it more flexible and more adaptable.

Kinetic parameters were estimated by calibrating the model with data from several operating AD systems. However, these constants need to be re-calibrated for mixed wastes.

Reasonable system performance results were obtained from the calculator. But further finetuning is required to arrive at more accurate results.

If only sales revenue from power produced is considered, and at the assumed costs of purchase/sold electricity, the CSTR and MPF systems would not have a positive economic return based on after-tax cash flow analysis.

Estimates of capital and operating costs need further investigation.

BC Ministry of Agriculture and Land – Gustav Rogstrand, Matt Dickson, Ann Eastman

BC Life Sciences (formerly BC Bioproducts Association)

Clean Energy Research Center, UBC

51

Acknowledgements

52

Thank you for your listening !

Questions ?