Journal of the Air & Waste Management Association

ISSN: 1096-2247 (Print) 2162-2906 (Online) Journal homepage: www.tandfonline.com/journals/uawm20

Biogas production from anaerobic digestion of

food waste and relevant air quality implications

Jeff Kuo & Jason Dow

To cite this article: Jeff Kuo & Jason Dow (2017) Biogas production from anaerobic digestion
of food waste and relevant air quality implications, Journal of the Air & Waste Management
Association, 67:9, 1000-1011, DOI: 10.1080/10962247.2017.1316326

To link to this article: https://doi.org/10.1080/10962247.2017.1316326

Published online: 22 Jun 2017.

Submit your article to this journal

Article views: 45715

View related articles

View Crossmark data

Citing articles: 42 View citing articles

Full Terms & Conditions of access and use can be found at

https://www.tandfonline.com/action/journalInformation?journalCode=uawm20
JOURNAL OF THE AIR & WASTE MANAGEMENT ASSOCIATION
2017, VOL. 67, NO. 9, 1000–1011
https://doi.org/10.1080/10962247.2017.1316326

TECHNICAL PAPER

Biogas production from anaerobic digestion of food waste and relevant air
quality implications
Jeff Kuoa and Jason Dowb
a
Department of Civil and Environmental Engineering, California State University, Fullerton, CA, USA; bCentral Marin Sanitation Agency, San
Rafael, CA, USA

ABSTRACT PAPER HISTORY

Biopower can diversify energy supply and improve energy resiliency. Increases in biopower production Received October 31, 2016
from sustainable biomass can provide many economic and environmental benefits. For example, Revised March 21, 2017
increasing biogas production through anaerobic digestion of food waste would increase the use of Accepted March 28, 2017
renewable fuels throughout California and add to its renewables portfolio. Although a biopower project
will produce renewable energy, the process of producing bioenergy should harmonize with the goal of
protecting public health. Meeting air emission requirements is paramount to the successful implemen-
tation of any biopower project. A case study was conducted by collecting field data from a wastewater
treatment plant that employs anaerobic codigestion of fats, oils, and grease (FOG), food waste, and
wastewater sludge, and also uses an internal combustion (IC) engine to generate biopower using the
biogas. This research project generated scientific information on (a) quality and quantity of biogas from
anaerobic codigestion of food waste and municipal wastewater sludge, (b) levels of contaminants in raw
biogas that may affect beneficial uses of the biogas, (c) removal of the contaminants by the biogas
conditioning systems, (d) emissions of NOx, SO2, CO, CO2, and methane, and (e) types and levels of air
toxics present in the exhausts of the IC engine fueled by the biogas. The information is valuable to those
who consider similar operations (i.e., co-digestion of food waste with municipal wastewater sludge and
power generation using the produced biogas) and to support rulemaking decisions with regards to air
quality issues for such applications.
Implications: Full-scale operation of anaerobic codigestion of food waste with municipal sludge is
viable, but it is still new. There is a lack of readily available scientific information on the quality of raw
biogas, as well as on potential emissions from power generation using this biogas. This research
developed scientific information with regard to quality and quantity of biogas from anaerobic co-
digestion of food waste and municipal wastewater sludge, as well as impacts on air quality from
biopower generation using this biogas. The need and performance of conditioning/pretreatment
systems for biopower generation were also assessed.

Introduction
Bioenergy is energy produced from biomass. It can be in the it represents 15.5% of total MSW generated. Due to
form of electricity (biopower), renewable gas (biomethane), difficulties in recovery/reuse of food waste, only 4.8%
or liquid transportation fuels (biofuels). Biopower can of the generated food waste was diverted from landfills
diversify energy supply and improve energy resiliency. and incinerators for recovery, mainly by composting
Increases in biopower production from sustainable biomass (EPA, 2014a). Food waste can be readily digested
can provide many economic and environmental benefits, under anaerobic conditions for capture of energy con-
including creation of green jobs, promotion of local eco- tent, and the residuals may be beneficially reused as
nomic stability, and reduction of water and air pollution fertilizer or soil amendment. Food waste has three
including greenhouse gas (GHG) emissions (O’Neill and times the methane (CH4) production potential of bio-
Nuffer, 2011; California Energy Commission [CEC], 2012). solids (376 vs. 120 m3 gas/ton) (EPA, 2014b). Yields
from anaerobic digestion can be as high as 3,200
standard cubic feet (scf), 90.6 m3, methane per ton
Energy recovery from food wastes
of raw food waste. Assuming an electricity cost of US
Total municipal solid waste (MSW) generation in the $0.10/kWh and an efficiency of 35% for converting
United States in 2012 was 251 million tons. Food biogas energy to electricity, energy in food waste
waste is the second largest component of MSW, and would be about $33/ton (Kraemer, 2012).

CONTACT Jeff Kuo [email protected] Department of Civil and Environmental Engineering, California State University, 800 N. State College
Boulevard, Fullerton, CA 92834, USA.
© 2017 A&WMA
 JOURNAL OF THE AIR & WASTE MANAGEMENT ASSOCIATION 1001

Anaerobic digestion of food wastes in wastewater Table 1. Comparison of anaerobic digestion of food waste and
treatment plants municipal wastewater sludge (source: EBMUD, 2008).
Parameter Units Food waste Wastewater sludge
Biogas is the main desirable product from anaerobic VS in feed (% of TS) % 86.3 77
digestion of organic carbon. Quality and quantity of bio- VS destruction % 73.8 38–57
CH4 content % 64 63
gas will be affected by many parameters, including pH, Methane production ft3/lb TS applied 6-8.5 5
temperature, feed composition, loading rate, mixing con- Note. 1 ft3/lb = 0.0623 m3/kg.
dition, reactor design, and residence time. Although
almost any organic material can be processed with anae-
mesophilic conditions and mean cell residence time
robic digestion, the level of digestibility is the key factor in
(MCRT) of 15 days. The data were extracted from a project
its successful application, if biogas production is the goal.
report by EBMUD (2008). As shown, the food waste tested
The more digestible the feed is, the higher is the gas yield
had a higher VS content than municipal wastewater sludge
potential. The biogas or methane yield is measured by the
(86.3 vs. 77%). Food waste is more digestible, as indicated
amount of biogas or methane that can be produced per
by the larger VS destruction (73.8 vs. 38–57%) after 15 days
unit mass of volatile solids (VS) contained in the feedstock
of anaerobic digestion, and has a larger methane formation
after a given amount of time under a given temperature
potential, 6–8.5 versus 5 ft3/lb (0.37–0.53 vs. 0.33 m3/kg)
(Banks et al., 2011; Zhang et al., 2007).
total solid (TS) applied. The methane concentrations in the
Many wastewater treatment plants (WWTPs) use anae-
biogas produced from these two types of feed are essentially
robic digestion to reduce the volume of biosolids before
the same, 64 versus 63%.
disposal and/or reuse. The amount of biogas produced by
existing WWTPs in California alone could fuel 125 MW of
power generation capacity. Co-digestion is the treatment of Biogas from anaerobic digestion
a mixture of at least two different substrates with an aim of
improving efficiency of anaerobic digestion. A recent study Using biogas in power generation (when compared to
by California Energy Commission (CEC) estimated that, fossil fuels) avoids additional GHG emissions (Razbani
using existing infrastructure, co-digesting fats, oil, and et al., 2011). With regard to beneficial uses of biogas
grease (FOG), food processing waste, and dairy waste at generated from anaerobic digestion, internal combustion
the existing WWTPs could increase the biogas yield poten- (IC) engines are the most common ones used in WWTPs
tial to 450 MW of capacity, representing 2,500 gigawatt (O’Neill, 2012). A flaring system and a boiler are often
hours (GWh) per year (Kulkarni, 2009). Food waste can needed to manage excess biogas during outage or main-
either be digested at facilities specifically designed for the tenance of the cogeneration system. Small WWTPs may
organic portion of MSW, or co-digested at WWTPs. need to supplement their biogas with natural gas to fulfill
WWTPs are ideal for accepting food waste diverted from the minimum fuel requirement (CH2M HILL, 2014).
landfills because the facilities are often located in urban Main components of biogas that are generated from
areas and have a short haul distance (i.e., less carbon anaerobic digestion are methane and carbon dioxide
footprint), have experiences in operating anaerobic diges- (CO2). However, biogas also contains other trace gases,
ters, and have existing infrastructure in place to capture moisture, particulate matters (PMs), and contaminants
biogas. In addition, large treatment facilities could use the such as volatile organic compounds (VOCs), sulfur com-
produced electricity and heat onsite (EPA, 2014b). It pounds, siloxanes, and ammonia. The composition of bio-
should be noted that preprocessing of food waste is often gas can be different in different plants and even different in
required because WWTP digesters can be damaged by a specific plant due to the differences in feed composition
highly fibrous material, metal, and plastic. Co-digesion and operating conditions of its anaerobic digesters. The
should also not exceed the design capacity of WWTP presence of several trace compounds in raw biogas pro-
digesters with regard to flow, solid loading, and biogas duced from anaerobic digestion may have adverse effects
handling (Kraemer, 2012). on beneficial uses. Removal of these trace compounds is
Many WWTPs are now in different stages of anaerobic often done through pretreatment (or conditioning). The
digestion of food waste (Institute for Local Self-Reliance most significant components targeted in biogas condition-
[ILSR], 2012; Kraemer, 2012; Moriarty, 2013; EPA 2014c). ing/pretreatment are hydrogen sulfide (H2S), siloxanes,
East Bay Municipal Utility District (EBMUD) in Oakland, moisture, PM, ammonia, and CO2. Hydrogen sulfide is a
CA, was the first large-scale WWTP in the nation to con- toxic product formed from sulfates and organic sulfur
vert postconsumer food scraps to energy through anaero- compounds in the feedstock under anaerobic conditions.
bic codigestion. Table 1 compares anaerobic digestion of During combustion, H2S will react with oxygen to form
food waste and municipal wastewater sludge under SO2, then sulfurous acid (H2SO3) and sulfuric acid
1002 J. KUO AND J. DOW

(H2SO4). These acids are corrosive to downstream equip- health and the environment. It is therefore important to
ment such as IC engines (Razbani et al., 2011). Stringent understand the emission profile of biogas from its creation
H2S limits are usually imposed by regulatory agencies. Gas to end use to ensure that the affected communities are not
treatments for H2S removal include adsorption, chemical further harmed. In addition, for example, many air districts
scrubbing, and biological scrubbing using biotrickling fil- in California are designated as nonattainment in regard to
ters (Huertas et al., 2011). Iron sponge adsorption is one of air quality standards of ozone and PMs. Emissions of
the most commonly used H2S removal systems, and alter- nitrogen oxides (NOx), precursors for ozone generation,
native iron oxide adsorbents such as SulfaTreat, Sulfur-Rite, and other compounds from stationary engines that utilize
and Sulfa-Bind are also being used (CH2M HILL, 2014). biogas are of concern. The NOx emission limits can be as
Siloxanes are often used in cosmetics, detergents, and low as 9 parts per million by volume (ppmv) in some air
building materials and are frequently found in household pollution control districts (Drake, 2011; U.S. Department
waste and wastewater. If siloxanes are present in the feed- of Agriculture [USDA]/EPA/Department of Energy
stock to the anaerobic digesters, the low-molecular-weight [DOE], 2015; Warner, 2009). Consequently, meeting air
siloxanes will volatilize into biogas. When this biogas is emission requirements is critical to the successful imple-
subsequently combusted in an IC engine, turbine, or boiler, mentation of any biopower project.
siloxanes will be converted into silicon dioxide and depos-
ited internally in the machine, exhaust manifolds, and
turbochargers, increasing wear and tear. Although food Objectives of this project
waste slurry should contain few or no siloxanes, they are Although anaerobic codigestion of food waste is a viable
often contained in biogas from codigestion with waste- process, full-scale operation of this process is still new.
water sludge. Activated carbon adsorption is currently the There is a lack of readily available scientific information
best available technology for removing siloxanes from bio- on the quality of raw biogas, as well as on potential
gas. Upstream removal of H2S and moisture are important emissions from power generation using this biogas. The
for optimal performance of the granulated activated carbon information is needed for determining conditioning
(GAC) absorbers. Silica gels are an alternative to GAC and requirements of raw biogas for beneficial uses, for select-
they are gaining acceptance as an option for siloxanes ing power-generating equipment, and for air quality
removal for their faster removal rates (CH2M HILL, 2014). permitting.
Nitrogen in the food waste typically enters the digesters The overall objectives of this research were (1) to
as organic nitrogen and a significant fraction of it is hydro- develop scientific information with regard to quality and
lyzed in the process, leaving the digesters as ammonia in quantity of biogas from anaerobic codigestion of food
the digestate. The level of ammonia concentration in the waste and municipal wastewater sludge, (2) to assess the
digestate of municipal WWTPs can be as high as 1,300 mg/ need and performance of conditioning/pretreatment sys-
L. Consequently, biogas also contains ammonia at a con- tems for biopower generation, and (3) to develop scien-
centration in equilibrium with that in the digestate. tific information with regard to impacts on air quality
Ammonia in the ambient air poses health risks and it can from biopower generation using this biogas.
be a precursor to airborne particles. Ammonia in the biogas
can react with water to form ammonium hydroxide
(NH4OH), which will corrode certain metals, such as alu- Project approaches and methods
minum and copper, making bearings more susceptible to
Treatment process trains
corrosion from ammonia (Razbani et al., 2011). It should
be noted that odorous compounds are often generated All field data of this research project were collected from
under anaerobic conditions. Proper odor control in areas the WWTP of Central Marin Sanitation Agency (CMSA),
of food waste processing and biogas generation and utiliza- San Rafael, California. The CMSA processes and disposes
tion may be needed. of approximately 10 million gallons per day (MGD), 0.44
There are many federal, state, and local regulations m3/sec, of wastewater and has treated in excess of 116
governing biogas production and uses in various aspects MGD (5.08 m3/sec) during peak rainfall periods. The
(air, water/wastewater and waste). As an example, the incoming raw wastewater goes through bar screens, aerated
“Permit Guidance for Anaerobic Digesters and Co-diges- grit chambers, primary clarifiers, biological treatment units
ters” provides the basic permitting framework and require- (i.e., biotowers + fine-bubble aeration tanks for activated
ments for anaerobic digestion projects in California sludge process), secondary clarifiers, chlorine contact
(California EPA, 2011). Although a biopower project will tanks, and then dechlorination before discharge into cen-
produce renewable energy, the process of producing bioe- tral San Francisco Bay through a 2-mile (3.2-km) outfall.
nergy should harmonize with the goal of protecting public The primary sludge (PS) and the thickened waste activated
 JOURNAL OF THE AIR & WASTE MANAGEMENT ASSOCIATION 1003

sludge (TWAS), which is thickened by rotary drum thick- system was replaced with a pump mixing system. The
eners (new), are fed to two mesophilic anaerobic digesters. floating cover of each digester was replaced with a two-
The biogas produced in the digesters is used to generate layer plastic membrane roof top with air in between to
electricity by an IC engine for treatment plant and facility regulate the pressure inside the digesters. The covers
uses and to heat plant process water. When digester biogas were replaced to increase available digester volume and
is not available, the IC engine switches to natural gas to avoid potential difficulties that could arise from using
(CMSA, 2013). floating covers (Kennedy/Jenks Consultants, 2008). The
The Central Marin Commercial Food-to-Energy two digesters at the project site are currently running in a
Program is a public–private partnership between CMSA mesophilic, single-stage, and continuous mode (Dow
and Marin Sanitary Service (MSS). The amount of food and Garbarino, 2013).
waste that could be collected, provided all the 500 food The biogas from the two anaerobic digesters goes
waste generators in the area participate into the program, through an H2S removal system and a siloxanes
is estimated to be up to 15 tons per day. The preconsumer removal system before being fed to the IC engine for
commercial food waste is collected and then transferred power generation. Figure 1 is the process flow diagram
to the MSS Transfer Station for processing, by uses of of biogas generation, conditioning, and utilization.
hoppers, belts, and magnet, and then transported to The raw biogas generated from anaerobic digestion
CMSA for further treatment (Dow and Garbarino, is first fed to a recently installed H2S-removal system
2013). On most days, the facility receives more than (Mi SWACO, Chesterfield, MO). The system consists
15,000 gallons/day (56.7 m3/day) of FOG, mostly from of two vessels (10 ft [3.05 m] diameter and 353 ft3 [10
restaurants in the MSS service area, and as of July 2016, m3] packing) filled with a synthetic blend of iron oxide
the facility recieves approximately 7.5 tons/day of food media (SulfaTreat 410 CHP; 4 × 16 mesh) operated in
waste up to 6 days per week. series and in a down-flow mode. The design gas velo-
The FOG and food waste are typically received in the city is 3.35 ft/min (1.02 m/min) and the design effluent
mornings. They are then mixed, ground, and recirculated H2S concentration is 15 ppmv with design influent H2S
in a storage tank for a couple of hours. The slurry is then concentration of 400 ppmv (maximum).
screened, by using a drum screen paddle mixer, to remove The effluent from the H2S removal system goes through
materials that are not readily digestible before being fed a condenser to reduce moisture content before being fed to
into the digesters (typically in late afternoons). The food two 2,500-lb (1,135-kg) activated carbon adsorbers (Model
waste slurry is fed to each digester on alternate days to be SAG 48V, Applied Filter Technologies, Snohomish, WA)
co-digested with PS and TWAS. for siloxanes removal. The two vessels (48 inch [1.2 m]
To upgrade for co-digestion of FOG and food waste, diameter × 72 inch [1.8 m] straight side, 45º bottom cone)
CMSA overhauled its 1985 anaerobic digestion system by operate in parallel and in an up-flow mode, and the design
installing new covers, mixers, biogas purification equip- effluent concentration is <100 ppbv of total siloxanes with
ment, and support systems (Creer, 2012). The gas mixing influent concentrations of 2 to 6 ppmv. CMSA currently

Figure 1. Process flow diagram of biogas generation, conditioning, and utilization.

1004 J. KUO AND J. DOW

budgets for one SulfaTreat media and two siloxanes media Method 2540G (TS and VS), SM 5220C,D (COD), SM
bed disposals per year (CMSA, 2014). 4500-H+ B (pH), EPA 170.1 (temperature), 4500-NH3 H
The effluent from the siloxanes removal system is fed to (ammonia), SM 2320 (alkalinity), and SM 5560 (VA).
the IC engine (Waukesha P48GLD, GE Power & Water,
Waukesha, WI) for cogeneration. The historic average of Biogas conditioning
run time on biogas is around 8 hr/day. With co-digestion A portable biogas analyzer, Gas Data GFM416 (Gas Data
of food waste and FOG, the runtime on biogas has Limited, Whitley, Coventry, UK) was acquired for this
increased to over 22 hr or longer (Dow, personal commu- project. The measurement ranges of the analyzer are CH4
nication, 2016). (0 to 100%), CO2 (0 to 100%), O2 (0 to 25%), and H2S (0 to
5,000 ppmv). Effluent samples of the biogas from the
Experimental approaches digesters, after the H2S removal system, and after the
siloxanes adsorbers were taken and analyzed for CH4,
The following experimental approaches were used in CO2, O2, and H2S twice daily by this portable biogas
this study: analyzer. The analyzer was routinely calibrated according
to manufacturer’s specifications. Selected samples were also
(1) To assess production rate and composition of analyzed for siloxanes (gas chromatography/flame ioniza-
raw biogas from anaerobic codigestion of food tion detection [GC/FID]) and VOCs (TO-15), as well as for
waste and municipal wastewater sludge. EPA Method 3C, “Determination of Carbon Dioxide,
(2) To evaluate removals of reduced sulfur com- Methane, Nitrogen, and Oxygen From Stationary
pounds and siloxanes from raw biogas by the gas Sources,” with testing by certified laboratories.
conditioning systems and to evaluate the energy
content of the biogas as well as the performance Emissions from the IC engine
and robustness of the conditioning systems. A portable emission analyzer, ECOM J2KN Pro Industrial
(3) To determine characteristics of emissions from OCNX-IR (ECOM America, Ltd., Gainesville, GA), was
the IC engine and the reliabilities and efficien- acquired for this project. The unit is equipped with sensors
cies of the system. for CH4, CO2, CO, NO2, NO, SO2, and O2. The emissions
were surveyed by the portable emission analyzer twice daily
(once when the engine was fueled by biogas and the once
Sampling plan and analytical methods when the engine was fueled by natural gas). The analyzer
was routinely calibrated according to manufacturer’s spe-
It should be noted that all the biogas and IC engine cifications. The range, accuracy, and resolution of the
emission samples collected in this study are for the pur- portable analyzer can be found in ECOM (2016).
pose of research only, and are not intended for use for Tests on IC engine emissions were also conducted by
regulatory compliance. Although samples were collected a Bay Area AQMD-certified source tester (Total Air
with care, the sampling approach might not meet all the Analysis, Carson, CA) on two separate days during the
requirements that are needed for compliance data. study period. For each test run, samples were taken and
Historical data of this facility prior to codigestion were analyzed for air toxics (i.e., formaldehyde, polycyclic
also included in the data analysis when appropriate, so aromatic hydrocarbons [PAHs], polychlorinated diben-
that comparisons could be made. zodioxins/furans [PCDD/F] and VOCs) by a certified
laboratory, Quantum Analytical Services (Carson, CA).
Biogas production and characteristics These IC engine emissions tests were conducted when
TS and VS of various components of the feed (i.e., PS, the engine was fueled by the biogas. The analytical
TWAS, FOG, and food waste) were sampled and analyzed methods used were CARB 428 (PCDD/F), CARB 429
on a daily basis. Selected samples were also analyzed for (PAHs), TO-14 (VOCs), and EPA 323 (formaldehyde).
chemical oxygen demand (COD). The loading rates of PS,
TWAS, FOG, and food waste were recorded and used to
calculate the mass loading rates to the digesters. Results and discussion
Operational temperature and pH of the digesters were
Characteristics of feed and digestate of anaerobic
also recorded. TS and VS of the digestate were analyzed
digestion
on a daily basis to facilitate the determination of VS
destruction. Concentrations of ammonia and volatile Characteristics of feed to the anaerobic digesters
acids (VA) and alkalinity of the digestate were also deter- The two digesters started to receive FOG in November
mined daily. Analytical methods used were Standard 2013 and food waste in February 2014. Due to
 JOURNAL OF THE AIR & WASTE MANAGEMENT ASSOCIATION 1005

considerable daily fluctuations in quantities and quality of kg/day), respectively. The contributions of PS, TWAS,
FOG and food waste during the 12-week sampling period FOG, and food waste to the TS and/or VS in the total
(08/08/14–11/07/14), determination of characteristics of feed were 49, 28, 15, and 9%, respectively. FOG and food
feed and digestate on a daily basis would not be mean- waste combined represents approximately 25% of TS or VS
ingful. Instead, analyses here were conducted by using the fed to the digesters. In other words, the addition of FOG
average values to delineate the trends. The typical hydrau- and food waste increased the VS loading to the digester by
lic residence time of anaerobic digestion was 36 days. one-third (25%/75%), when compared to that without the
Table 2 provides some statistics on characteristics and addition. This implies the biogas formation would increase
flow rates of the feed streams to the anaerobic digesters. at least 33% with codigestion since the VS of FOG and food
(The raw data of this research project can be found in Kuo waste are more readily digestible than that in the municipal
[2015].) As shown, the average TS values for PS, TWAS, wastewater sludge.
FOG, and food waste were 4.4 ± 0.6%, 4.5 ± 0.7%, 3.1 ±
2.1%, and 20.6 ± 3.3%, respectively. The data indicate that Characteristics of the digestate
the TS concentrations for PS, TWAS, and food waste are Table 4 provides some statistics on characteristics of the
relatively consistent, while those for FOG vary consider- digestate. Assuming the flow rates of the feed and the
ably. The average percentages of VS in the TS were 84 ± digester effluent are the same, the calculated VS destruction
2%, 83 ± 1%, 91 ± 4%, and 90 ± 3% for PS, TWAS, FOG, is 11,460 lb (5,180 kg) VS/day and the VS destruction
and food waste, respectively. As expected, the organic con- efficiency is 64.9% (lb VS destructed/lb VS applied). The
tents of FOG and food waste (91% and 90%) are higher destruction efficiency is in line with the data in the litera-
than those of PS and TWAS (84% and 83%). Samples were ture, as shown in Table 1 (EBMUD, 2008). The average
also taken from the combined feed stream of FOG and values of pH, temperature, alkalinity, and NH4+-N were 7.2
food waste; the TS, percentage of VS, and COD were 5.5 ± ± 0.0, 99.7 ± 0.3ºF (or 37.6 ± 0.2ºC), 4,853 ± 179 mg/L, and
5.7%, 88.6 ± 7.2%, and 39,900 ± 28,600 mg/L, respectively. 1,137 ± 83 mg/L, respectively. As shown, all these opera-
The average TS and VS values are comparable to the tional parameters are in narrow ranges, and this implies the
corresponding flow-rate weighted average values of the anaerobic digesters were being operated under stable con-
FOG and food waste. ditions. The ammonium concentrations were around 1,150
By using the average values of TS, VS, and loading rate mg/L, and that did not seem to be inhibitory to biological
of each stream (i.e., PS, TWAS, FOG, and food waste), the activities, as evidenced by increases in biogas production.
corresponding values of the total feed stream to the anae-
robic digesters were calculated and are shown in Table 3.
The average total flow rate to the digesters was 57,300 Biogas production and conditioning
gallons/day (217 m3/day), which contained 4.5% TS, and Production and characteristics of biogas generated
the VS percentage was 85.1%. The average mass loadings of Table 5 provides some statistics on characteristics of the
TS and VS were 21,500 and 18,300 lb/day (9,760 and 8,310 biogas from the digesters. The concentrations of CH4, CO2,

Table 2. Characteristics and flow rates of feed streams to the anaerobic digesters.
PS TWAS FOG Food Waste
TS (%) VS (%) Rate (gal/day) TS (%) VS (%) Rate (gal/day) TS (%) VS (%) Rate (gal/day) TS (%) VS (%) Rate (ton/day)
Count 59 59 65 59 59 65 60 60 58 70 70 71
Maximum 7.8 88 42,072 5.3 92 26,685 11.0 98 20,500 29.3 94 8.6
Minimum 3.5 80 17,630 3.8 80 8,930 0.4 77 4,000 12.9 72 1.7
Median 4.3 84 28,069 4.6 83 16,016 2.4 92 10,500 20.4 91 3.6
Average 4.4 84 28,588 4.6 83 16,098 3.1 91 11,543 20.6 90 4.1
Std. dev. 0.6 2 4,769 0.7 11 2,967 2.1 4 4,316 3.3 3 1.9
Note. 1 gallon/day = 3.785 L/day; 1 ton/day = 2,000 lb/day = 908 kg/day.

Table 3. Characteristics and loading rates of total feed to the anaerobic digesters.
PS TWAS FOG Food waste Total daily feed
Loading rate (gallons/day) 28,588 16,098 11,543 1,077 57,305
TS (%) 4.4 4.6 3.1 20.6 4.5
VS (%) 83.7 82.7 91.5 90.4 85.1
TS (lb/day) 10,543 6,155 2,943 1,854 21,495
VS (lb/day) 8,828 5,089 2,692 1,676 18,285
Percent in TS of total feed 49.0 27.8 14.7 9.2 100.0
Percent in VS of total feed 48.3 27.8 14.7 9.2 100.0
Note. 1 gallon/day = 3.785 L/day; 1 lb/day = 0.454 kg/day.
1006 J. KUO AND J. DOW

Table 4. Characteristics of the digestate. than that for the VS loading rate (38%) supports the
TS VS Temperature VA Alkalinity NH4+-N argument that volatile solids in FOG and food waste are
(%) (%) pH (°F) (mg/L) (mg/L) (mg/L)
more readily biodegradable than those in the municipal
Count 59 59 59 60 60 60 58
Maximum 2.2 70 7.3 100.7 162 5,300 1,322 wastewater sludge.
Minimum 1.9 63 7.2 99.3 90 4,600 1,008 Also shown in Table 5, the average CH4 concentrations
Median 2.0 68 7.2 99.7 122 4,850 1,119
Average 2.0 67 7.2 99.7 124 4,853 1,137 in the raw biogas, effluent from the H2S adsorbers, and
Std. dev. 0.1 2 0.0 0.3 18 179 83 effluent from the siloxanes adsorbers were 62.6 ± 0.7%,
Note. °C = (°F – 32) × (5/9). 62.8 ± 0.7%, and 63.6 ± 0.5%, respectively. These values
are essentially the same, which implies that these two
biogas conditioning systems have no, or insignificant,
O2, and H2S in the raw biogas were 62.6 ± 0.7%, 36.1 ±
effects on methane concentrations of the biogas. On the
0.7%, 0.0 ± 0.0%, and 127 ± 75 ppmv, respectively. With
other hand, the H2S concentrations dropped from 127 ±
the average methane concentration of 62.6%, the heating
75 to 7 ± 13 ppmv by the H2S removal system. The system
value of the raw biogas is slightly greater than 600 BTU/ft3
appeared to meet the design specification. Hydrogen sul-
(22,340 kJ/m3).
fide was not detected in the effluent of the siloxanes
The daily biogas extraction rate, not necessarily the
adsorbers, plausibly due to the additional removal of
biogas generation rate, depended mainly on two operating
H2S by GAC in the adsorbers.
factors. The first was to keep the level of the dome and the
Grab samples of raw biogas, biogas in the effluent of
pressure in the dome within the recommended ranges of
the H2S adsorbers, and biogas in the effluent of the
the manufacturer; the second was to make sure that there
siloxanes adsorbers were analyzed for EPA Method 3C,
was enough gas to run the IC engine during the peak power
and the data served as a quality assurance and quality
window, 12 to 6 pm. As shown in Table 5, the daily biogas
control (QA/QC) check for the measurements of the
flow rate was 212,800 ft3/day (6,030 m3/day). The corre-
portable biogas analyzer. The results are summarized
sponding biogas generation rate is 10.2 ft3 biogas/lb (0.64
in Table 6 and the values between the portable biogas
m3 biogas/kg) TS applied or 6.4 ft3 CH4/lb (0.40 m3 CH4/
analyzer and the certified lab are comparable.
kg) TS applied, which is within the range of 6 to 8.5 ft3
CH4/lb (0.37 to 0.53 m3 CH4/kg) TS applied for food waste
in literature, while the corresponding value for the Performance of the hydrogen sulfide removal system
municipal wastewater sludge is 5 ft3 CH4/lb (0.31 m3 Grab samples were also analyzed for reduced sulfur com-
CH4/kg) TS applied (EBMUD, 2008). On the basis of VS pounds. Out of 22 reduced sulfur compounds, H2S and
destruction, the biogas generation rate was 18.5 ft3 biogas/ n-propyl mercaptan were the only two above the detection
lb (1.15 m3 biogas/kg) VS destroyed, or 11.6 ft3 CH4/lb limit of 0.065 ppmv (Table 7). The concentration of total
(0.72 m3 CH4/kg) VS destroyed. The average H2S concen- reduced sulfur compounds as H2S in the raw biogas was
tration in the raw biogas was 127 ± 75 ppmv. 164 ppmv, and it was reduced to 14.6 ppmv by two H2S
During the similar period in 2011 (08/18/11 to 11/07/ adsorbers. It was further reduced to 10.4 ppmv by the
11) when the anaerobic digesters only received PS and siloxanes adsorbers. This serves as a QA/QC check for the
TWAS, the average biogas generation rate was 131,800 ± H2S measurements by the portable biogas analyzer and the
26,800 ft3 (3,730 ± 760 m3)/day. With co-digestion of results are comparable.
FOG, food waste, PS, and TWAS, the daily biogas yield
has increased from 131,800 to 212,800 ft3 (3,730 to 6,030 Performance of the siloxanes removal system
m3)/day, a 61% increase. On a mass loading basis, the Samples of raw biogas, biogas in the effluent of the H2S
biogas yield has increased from 8.75 to 12.1 ft3/lb (0.55 to adsorbers, and biogas in the effluent of the siloxanes adsor-
0.75 m3/kg) VS entering the digesters, a 38% increase. The bers were grabbed on 10/08/14 and 10/29/14 and analyzed
fact that the increase in biogas production (61%) is larger by a certified lab for siloxanes: hexamethyldisiloxane (L2),

Table 5. Daily biogas production rate and composition.

Raw biogas Post H2S adsorbers Post siloxanes adsorbers
Date Biogas flow rate (ft3/day) CH4 (%) CO2 (%) H2S (ppm) CH4 (%) CO2 (%) H2S (ppm) CH4 (%) CO2 (%) H2S (ppm)
Count 52 52 52 52 52 52 52 45 45 45
Maximum 330,079 64.3 37.2 323 64.8 37.5 43 64.8 37.0 0
Minimum 112,098 61.3 33.2 0 61.3 34.4 0 61.9 34.9 0
Median 199,827 62.7 36.4 133 62.8 36.3 0 63.6 36.1 0
Average 212,811 62.6 36.1 127 62.8 36.3 7 63.6 36.0 0
Std. dev. 46,997 0.7 0.7 75 0.7 0.6 13 0.5 0.4 0
Note. 1 ft3/day = 0.0283 m3/day.
 JOURNAL OF THE AIR & WASTE MANAGEMENT ASSOCIATION 1007

Table 6. Results of EPA Method 3C testing (concentrations in % heptane), two ketones (2-butanone and 4-methyl-2-
by volume). pentanone), four aromatics (benzene, toluene, ethyl-
Compound Raw biogas Post H2S adsorbers Post siloxanes adsorbers benzene, and xylenes) and tetrahydrofuran were
H2 <1.3 <1.3 <1.3
O2 <0.1 <0.1 <0.1
detected. Table 9 tabulates the concentrations of those
N2 0.7 0.6 0.4 compounds detected in the raw biogas, effluent of the
CO <0.1 <0.1 <0.1
CO2 34.9 34.7 34.4 H2S adsorbers, and effluent of the siloxanes adsorbers.
CH4 64.3 64.6 65.1

Table 7. Removal of the sulfur compounds in biogas by the IC engine emissions

conditioning systems (concentrations in ppmv)
Raw Post H2S Post siloxanes
Historic source test results on IC engine emissions
Compound biogas adsorbers adsorbers Since the biogas generated at the site is not sufficient for
Hydrogen sulfide 163 14.0 10.0 around-the-clock power generation, the engine is fueled
n-Propyl mercaptan 0.824 0.604 0.367
Total reduced sulfur 164 14.6 10.4 by natural gas when biogas is not available. The IC engine
as H2S has been source-tested by Bay Area Air Quality
Management District (BAAQMD) annually to check for
compliance. For each test, the IC engine was fueled by
hexamethylcyclotrisiloxane (D3), octamethyltrisiloxane biogas as well as by natural gas. The results from these
(L3), octamethylcyclotetrasiloxane (D4), decamethyltetra- official source tests provide a valuable opportunity to
siloxane (L4), decamethylcyclopentasiloxane (D5), and compare the emissions from the same engine on the
dodecamethylpentasiloxane (L5). same day by using biogas versus natural gas.
As shown in Table 8, only D3, D4, and D5 were detected Table 10 provides some statistics of the annual source
in the raw biogas (the detection limit is 13.0 ppbv). The test results from 2008 to 2014 and the IC engine has been
siloxanes adsorbers reduced the D3 and D5 concentrations in compliance. As shown, the average emissions from
to below or close to the detection limit, and the removal of natural gas (NG)-fueled and biogas-fueled operations
D4 was also 80% or better. However, it should be noted are: NOx at 15% O2 (38 ± 19 vs. 37 ± 20 ppm); CH4
that the concentrations of D3 in the effluent of the H2S (692 ± 83 vs. 1,065 ± 224 ppm); nonmethane organic
adsorbers were higher than those in the raw biogas on both carbon (NMOC) as C1 (45 ± 33 vs. 18 ± 6 ppm); total
days (295 vs. <13.0 and 270 vs. 204 ppbv). In addition, the organic carbon (TOC) as C1 (735 ± 65 vs. 1,075 ± 227
concentrations of L2 in the effluent of the siloxanes adsor- ppm); CO2 (7.5 ± 0.5 vs. 12.8 ± 1.6%); CO at 15% O2 (122
bers were higher than those in the effluent of the H2S ± 9 vs. 142 ± 14 ppm); O2 (7.9 ± 0.4 vs. 6.9 ± 0.7%); and
adsorbers in both days (84 vs. <13.0 and 63.4 vs. 13.0 SO2 (4 ± 3 vs. 7 ± 5 ppm).
ppbv). The causes for the increases were not identified. The NOx concentrations in the exhausts using NG and
One plausible reason is that these samples were not taken biogas are essentially the same (38 vs. 37 ppmv). With
in a synchronized manner. The total effluent concentra- regard to emissions of organic compounds, NG-fueled
tions of siloxanes were 99.1 ppbv on 10/08/14 and 117.9 operations emit less CH4 (692 vs. 1,065 ppmv), but higher
ppbv on 10/29/14, which are either at or slightly above the NMOC (45 vs. 18 ppmv). The average CO2 emission from
design specification of 100 ppbv. using biogas is higher (12.8 vs. 7.5%), probably due to the
These samples were also analyzed for VOCs using higher CO2 concentrations in the biogas. The average CO
EPA Method TO-15. Out of a long list of compounds concentration from biogas-fueled combustion was higher
analyzed for, only one alkene (propene), four alkanes (142 vs. 122 ppmv). The average SO2 concentration from
(cyclohexane, 2,2,4-trimethylpentane, hexane, and biogas-fueled combustion was also higher (7 vs. 4 ppmv),

Table 8. Removal of siloxanes in biogas by the conditioning systems (concentrations in ppbv).

10/8/2014 10/29/2014
Raw Post H2S Post siloxanes Raw Post H2S Post siloxanes
biogas adsorbers adsorbers biogas adsorbers adsorbers
Hexamethyldisiloxane (L2) <13.0 <13.0 84 <13.0 <13.0 63.4
Hexamethylcyclotrisiloxane (D3) <13.0 295 <13.0 204 270 <13.0
Octamethyltrisiloxane (L3) <13.0 <13.0 <13.0 <13.0 <13.0 <13.0
Octamethylcyclotetrasiloxane (D4) 161 225 15.1 202 202 41.6
Decamethyltetrasiloxane (L4) <13.0 <13.0 <13.0 <13.0 <13.0 <13.0
Decamethylcyclopentasiloxane (D5) 1160 570 <13.0 523 334 13.9
Dodecamethylpentasiloxane (L5) <13.0 <13.0 <13.0 <13.0 <13.0 <13.0
1008 J. KUO AND J. DOW

Table 9. Removal of VOCs in biogas by the conditioning systems (concentrations in ppbv).

10/8/2014 10/29/2014
Compound Raw biogas Post H2S adsorber Post siloxanes adsorber Raw biogas Post H2S adsorber Post siloxanes adsorber
Propene 1,730 1,700 1,590 1,150 1,180 1,140
Cyclohexane <6.5 <6.5 9.9 <6.5 <6.5 <6.5
2,2,4-Trimethylpentane <6.5 7.2 84.4 <6.5 <6.5 49.9
Hexane 29.8 29.7 110 17.1 16.4 52.4
Heptane 130 106 238 53.2 54.1 140
2-Butanone 31.2 35.8 39.2 <13.0 <13.0 <13.0
4-Methyl-2-pentanone <6.5 <6.5 13.7 <6.5 <6.5 10.6
Benzene 20.5 21.4 72.1 17.4 17.9 35.3
Toluene 735 690 1,860 1,430 1,430 2,130
Ethylbenzene 19.4 17.0 55.7 18.4 20.9 41.8
m- and p-Xylenes 18.4 16.0 20.8 14.7 15.7 22.7
o-Xylene 6.7 <6.5 <6.5 <6.5 <6.5 <6.5
Tetrahydrofuran 88.5 <6.5 10.3 <6.5 <6.5 <6.5

Table 10. Statistics of IC engine emission data using natural gas versus biogas (official source test results from 2008–2014).
Natural gas Biogas
Maximum Minimum Median Average Std. dev. Maximum Minimum Median Average Std. dev.
Output (kW) 650 538 590 601 42 649 550 620 612 30
Flow rate (SCFM) 2,140 1,390 1,720 1,717 227 1,810 1,340 1,590 1,583 176
CO (ppm) 292 246 263 269 17 381 286 329 334 28
CO (ppm), converted to 15% O2 138 106 121 122 9 157 114 142 142 14
CO (gm/Hp-hr) 1.56 0.78 1.14 1.14 0.21 1.58 0.87 1.28 1.29 0.21
NO (ppm) 129 20 29 49 37 118 11 46 55 42
NO2 (ppm) 49 24 33 36 8 43 18 36 33 8
NOx (ppm), converted to 15% O2 77 24 28 38 19 66 13 38 37 20
NOx (g/Hp-hr) 0.93 0.37 0.45 0.55 0.20 1.02 0.21 0.51 0.54 0.30
Methane (ppm) 799 561 686 692 83 1,400 762 1,070 1,065 224
NMOC as C1 (ppm) 90 10 34 45 33 26 10 18 18 6
NMOC as C1 (g/Hp-hr) 0.21 0.02 0.08 0.10 0.07 0.07 0.02 0.04 0.04 0.02
TOC as C1 (ppm) 799 639 760 735 65 1,418 778 1,070 1,075 227
CO2 (%) 8.3 6.8 7.3 7.5 0.5 15.1 10.7 12.7 12.8 1.6
Oxygen (%) 8.5 7.2 8 7.9 0.4 8.2 6 6.8 6.9 0.7
SO2 (ppm) 11 2 2 4 3 16 2 6 7 5
SO2 (g/Hp-hr) 0.08 0.01 0.02 0.03 0.03 0.15 0.02 0.05 0.06 0.04
Estimated heat input (MMBTU/d) 191 191 191 191 0
Heating value of biogas (BTU/scf) 634 424 560 542 65
Note. 1 CFM = 0.0283 m3/min; 1 g/Hp-hr = = 1.34 g/kWh; 1 BTU/day = 1,055 J/day; 1 BTU/ft3 = 37.24 kJ/m3.

probably due to the presence of reduced sulfur compounds similar to those measured by the portable emission ana-
in the biogas. lyzer (Table 11). If the stringent 2016 regulations in the
jurisdiction area of South Coast Air Quality
Daily monitoring of IC engine emissions Management District (SCAQMD) in southern
Table 11 provides some statistics of the daily IC engine California were enforced on this IC engine, the CO con-
emission data. As shown, the average emissions from NG- centration (at 15% O2) of 157 ppmv would be less than
fueled and biogas-fueled operations are: NO2 (30 ± 3 vs. 23 the 250 ppmv standard; however, the NOx concentration
± 4 ppm); NO (17 ± 4 vs. 10 ± 3 ppm); CH4 (1,200 ± 170 vs. (at 15% O2) would be higher than the 11 ppmv standard
1,680 ± 200 ppm); CO2 (7.2 ± 0.1 vs. 7.7 ± 0.1%); CO (122 (SCAQMD, 2012). It should be noted that the NOx con-
± 9 vs. 142 ± 14 ppm); O2 (8.0 ± 0.1 vs. 7.2 ± 0.1%); and SO2 centratons from this IC engine, using natural gas, are also
(0 ± 0 vs. 0 ± 0 ppm). The data from daily monitoring are higher than 11 ppmv, as indicated by the official source
comparable to those of the source tests. test results from 2008 to 2014 (see Table 10).
With regard to air toxics, the formaldehyde concentra-
Toxics testing on the IC engine emissions tions were 31.7 ppbv (10/29/14) and 32.6 ppbv (10/30/14)
Tests on IC engine emissions, when the engine was fueled in the IC engine exhausts. For the VOC analysis (EPA
by biogas, were also conducted by a BAAQMD-certified Method TO-14), two aromatics (benzene and toluene)
source tester on 10/29/14 and 10/30/14. and 10 halogenated organic compounds (1,1,1-trichlor-
Table 12 tabulates the emission test results from the oethane [TCA], 1,2-dichloropropane, bromomethane,
source tester and the data are similar to those of the carbon tetrachloride, chloroethane, chloroform, cis-1,2-
official source tests in the past 7 years (Table 10). The dichloroethylene [DCE], trichloroethylene [TCE], tri-
concentrations of CH4, CO2, CO, NOx, and O2 are also chlorotrifluoroethane, and vinyl chloride) were detected
 JOURNAL OF THE AIR & WASTE MANAGEMENT ASSOCIATION 1009

Table 11. Daily IC engine emission data.

Natural gas Biogas
CH4 CO2 CO NO2 NO SO2 O2 CH4 CO2 CO NO2 NO SO2 O2
(%) (%) (ppm) (ppm) (ppm) (ppm) (%) (%) (%) (ppm) (ppm) (ppm) (ppm) (%)
Count 38 38 38 38 38 38 38 54 54 54 54 54 54 54
Maximum 0.195 7.7 378 37 28 0 8.2 0.214 7.8 385 30 18 0 7.3
Minimum 0.092 7.1 224 21 10 0 7.2 0.127 7.6 160 17 4 0 7.0
Median 0.119 7.2 312.5 30 17 0 8.0 0.170 7.7 367 22.5 10.5 0 7.2
Average 0.120 7.2 312 30 17 0 8.0 0.168 7.7 360 23 10 0 7.2
Std. dev. 0.017 0.1 19 3 4 0 0.1 0.020 0.0 36 4 3 0 0.1

Table 12. IC engine emissions from tests conducted by the Table 13. VOCs in the IC engine exhausts from tests conducted
source tester. by the source tester (concentrations in ppbv).
10/29/2014 10/30/2014 10/29/2014 10/30/2014
Output (kW) 613 615 Benzene 33 30
Measured flow rate (SDCFM) 1,886 1,898 Toluene 12 11
Calculated flow rate (SDCFM) 1,972 2,190 1,1,1-Trichloroethane 0.1 0.1
CO (ppm) 362 361 1,2-Dichloropropane 1.4 0.9
CO (ppm), converted to 15% O2 157 157 Bromomethane 648 1,027
CO (g/Hp-hr) 1.63 1.81 Carbon tetrachloride 0.1 0.1
NOx (ppm) 54.8 51.4 Chloroethane 111 102
NOx (ppm), converted to 15% O2 23.8 22.3 Chloroform 22 16
NOx (g/Hp-hr) 0.33 0.35 cis-1,2-Dichloroethene 17 23
Methane (ppm) 1,484 1,708 Trichloroethene 0.2 0.4
Methane (ppm), converted to 15% O2 643 741 Trichlorotrifluoroethane 0.5 <0.3
TOC as C1 (ppm) 1,484 1,708 Vinyl chloride 167 604
CO2 (%) 11.3 11.3
O2 (%) 7.23 7.24
Note. 1 CFM = 0.0283 m3/min; 1 g/Hp-hr = = 1.34 g/kWh.
Table 14. Dioxins/furans in the IC engine exhausts from tests
conducted by the source tester.
(see Table 13). If the stringent 2016 SCAQMD regulations 10/29/2014 10/30/2014
were enforced on this IC engine, the total VOC concen- pg/sample mg/dscm pg/sample mg/dscm
OCDD 31.2 1.38E-08 11.7 5.07E-09
trations, which are at the level of a couple of parts per Total HpCDD 7.77 3.42E-09 <9.55 <4.14E-09
million by volume or less, would be less than the 30 ppmv Total TCDF 44.2 1.95E-08 <8.72 <3.78E-09
Total PeCDF 39.1 1.72E-08 <7.98 <3.46E-09
standard (SCAQMD, 2012). Total HxCDF 7.65 3.37E-09 <3.95 <1.71E-09
From using CARB Method 428, octachlorodibenzo-
dioxane (OCDD) was the only single PCDD/F species
detected. In one of the two sampling events the masses PAH species were detected in the IC engine exhausts
of total heptachlorodibenzodioxine (HpCDD), tetra- and naphthalene has the highest average concentration,
chlorodibenzofuran (TCDF), pentachlorodibenzofuran at 6.1 × 10−3 mg/dscm (no specific discharge limits on
(PeCDF), and hexachlorodibenzofuran (HxCDF) were these compounds were found from a literature search).
also above the detection limits. Table 14 tabulates the Two biogas samples (one from each testing day)
masses of the detected compounds in the collected were also analyzed for higher heating values (HHV)
samples and the corresponding concentrations (mg/ using ASTM 1945-03. The gross heating value is 639
dry standard cubic meter), calculated using the mass BTU/ft3 (23,800 kJ/m3) and the net heating value is 575
and the total sample volume (80.11 and 81.49 ft3 [2.27 BTU/ft3 (21,400 kJ/m3) on average.
and 2.31 m3] for 10/29/14 and 10/30/14, respectively).
As shown, the concentrations are relatively low, at 1.95
Conclusion
× 10−8 mg/dscm or less. (In addition to concentrations,
the discharge limits of these compounds from an IC The findings from this study include the following:
engine typically depend on discharge flow rate, stack
configuration, meterological conditions, and popula- (1) With FOG and food waste making up approxi-
tion density and distribution of affected communities. mately 25% of TS or VS loading to the anaero-
No specific discharge limits on these compounds were bic digesters, the digesters are being operated
found from a literature search.) under stable conditions.
The exhausts were also collected and analyzed for (2) The biogas production rate is 18.5 ft3 biogas/lb
PAHs using CARB Method 429. Table 15 tabulates the (1.15 m3 biogas/kg) VS destroyed. With 33%
masses of the detected compounds in the collected percent more VS loading from FOG and food
samples and the corresponding concentrations. Ten waste, the daily biogas production is 60% more.
1010 J. KUO AND J. DOW

Table 15. PAHs in the IC engine exhausts from tests conducted additional emission control may be needed to
by the source tester. meet the low NOx (at 15% O2) limit of 11
10/29/2014 10/30/2014 ppmv.
ng/sample mg/dscm ng/sample mg/dscm
Naphthalene 13,300 5.86E-03 14,600 6.33E-03
2-Methylnaphthalene 8,240 3.63E-03 11,100 4.81E-03
Acenaphthylene 773 3.41E-04 815 3.53E-04
Acenaphthene 357 1.57E-04 458 1.98E-04
Fluorene 698 3.08E-04 352 1.53E-04
Acknowledgment
Phenanthrene 2,470 1.09E-03 1,750 7.58E-04
Anthracene 128 5.64E-05 <20 8.67E-06 The research team expresses its gratitude toward the commis-
Fluoranthene 246 1.08E-04 142 6.15E-05 sion contract managers, Marla Mueller and Simone Brant, and
Pyrene 193 8.51E-05 108 4.68E-05 Guido Franco and Yu Hou of the California Energy
Chrysene 85.1 3.75E-05 47.7 2.07E-05
Commission (CEC), as well as Robert Dole (recently retired)
and many other staff members of the CMSA for providing great
guidance and assisting the team in many ways. The research
(3) H2S is the dominant reduced sulfur compound team also expresses gratitude toward the Technical Advisory
in the raw biogas, while n-propyl mercaptan is Committee (TAC) members for providing valuable technical
another one detected. The on-site treatment input and suggestions for this project. The TAC members are
system is capable of reducing the H2S concen- (in alphabetical order) Rizaldo Aldas (CEC), Chris Berch
tration below 15 ppmv. (Inland Empire Utility Agency), Ryann Bonner (Environ
Strategy), Robert Cole (CMSA), Jason Dow (CMSA), Steven
(4) With regards to siloxanes, only D3, D4, and D5 Fan (City of Los Angles), Jacques Franco (CalRecyle), Robert
were detected in the raw biogas. The on-site acti- Gilles (San Joaquin Valley APCD), Ken Kumar (Energy
vated carbon adsorbers reduced the D3 and D5 Environmental Solutions), Tung Le (CARB), Angus
concentrations to below or close to the detection MacPherson (CARB), Gary O’Neill (CEC), Lisa Van de Water
limit, and the removal of D4 was 80% or better. (San Joaquin Valley APCD), and Robert Williams (UC Davis).
(5) In the raw biogas samples, only one alkene, four
alkanes, two ketones, four aromatics, and tetra- Funding
hydrofuran were detected using EPA Method
This project was funded by the California Energy
TO-15 analysis. With regard to air toxics in the
Commission under the Public Interest Energy Research
IC engine exhausts, formaldehyde concentration (PIER) Program (Agreement Number 500-11-030).
was 32 ppbv. Two aromatics and 10 halogenated
organic compounds were detected using EPA
Method TO-14. About the authors
(6) The IC engine at the site used both NG and Jeff Kuo is a professor at the Department of Civil and
biogas on a daily basis. The NOx concentrations Environmental Engineering, California State University,
(at15% O2) in the exhausts using NG and biogas Fullerton (CSUF).
were essentially the same (38 vs. 37 ppmv). NG-
Jason Dow is the General Manager of Central Marin
fueled operations emitted less CH4 (692 vs. 1,065 Sanitation Agency (CMSA), San Rafael, CA.
ppmv), but higher nonmethane organic carbon
(NMOC) (45 vs. 18 ppmv). Both the average CO2
and CO (at 15% O2) concentrations from using References
biogas were higher than those using NG (12.8 vs.
Banks, C.J., M. Chesshire, S. Heaven, and R. Arnold. 2011.
7.5% for CO2 and 142 vs. 122 ppmv for CO, Anaerobic digestion of source-segregated domestic food
respectively). The average SO2 concentration waste: Performance assessment by mass and energy bal-
from biogas-fueled combustion was also higher ance, Bioresource Technol. 102:612–20. doi:10.1016/j.
than that using NG (7 vs. 4 ppmv). biortech.2010.08.005
(7) OCDD was the only single PCDD/F species California Energy Commission. 2012. Addressing air quality issues
for biopower, a request for research proposals from California
detected in the IC engine exhausts, at an aver- public universities. California Energy Commission, January.
age concentration of 9.4 × 10−9 mg/dscm. http://lawr.ucdavis.edu/pdf/CFP/305_INFORMAL-RFP-
(8) Ten PAH species were detected in the IC Addressing-AirQuality-Issues-for-Biomass-final-draft.pdf
engine exhausts with naphthalene having the California Environmental Protection Agency. 2011. Permit
highest concentration, at 6.1 × 10−3 mg/dscm. guidance for anaerobic digesters and co-digesters,
Version 2 California EPA, December 22. http://www.
(9) The IC engine fueled by biogas can meet strin-
calepa.ca.gov/digester/Documents/GuideDigester.pdf
gent emission limits for CO at 15% O2 (250 Central Marin Sanitation Agency. 2013. Adopted operating
ppmv) and VOCs (30 ppmv). However, and capital budget, fiscal year 2013–2014. Central Marin
 JOURNAL OF THE AIR & WASTE MANAGEMENT ASSOCIATION 1011

Sanitation Agency. http://mygfoa.net/budgetapp/2013/ Commission, CEC-500-2015-037. http://www.energy.ca.

CentralMarinSanitationAgencyCA.pdf gov/2015publications/CEC-500-2015-037/CEC-500-2015-
Central Marin Sanitation Agency. 2014. Proposed operating 037.pdf
and capital budget, fiscal year 2014–2015. Central Marin Moriarty, K. 2013. Feasibility study of anaerobic digestion of
Sanitation Agency, http://www.cmsa.us/assets/meetings/ food waste in St. Bernard, Louisiana. Report prepared for
collateral/FY15_Budget_-_Proposed__website_.pdf National Renewable Energy Laboratory (NREL) of
CH2M HILL. 2014. Las Gallinas Valley Sanitation District— Department of Energy, NREL/TP-7A30-57082. http://
Biogas utilization technologies evaluation. Final technical www.nrel.gov/docs/fy13osti/57082.pdf
memorandum prepared by Dan Robillard of CH2M HILL O’Neill, G. 2012. 2012 Bioenergy action plan, California
for Las Gallinas Valley Sanitation District, April 19. http:// Energy Commission, Efficiency and Renewables Division.
www.lgvsd.org/wp-content/uploads/LGVSD-Biogas- http://resources.ca.gov/docs/2012_Bioenergy_Action_Plan.
Utilization-Evaluation.pdf pdf
Creer, D. 2012. Energy efficiency and biogas generation at O’Neill, G., and J. Nuffer. 2011. 2011 Bioenergy action plan,
wastewater plants. BioCycle 53(7):37. http://www.biocycle. California Energy Commission, Efficiency and Renewable
net/2012/07/17/energy-efficiency-and-biogas-generation- Division. Publication number CEC-300-2011-001-CTF.
at-wastewater-plants http://www.energy.ca.gov/2011publications/CEC-300-
Dow, J., and P. Garbarino. 2013. Central Marin Commercial 2011-001/CEC-300-2011-001-CTF.PDF
Food-to-Energy (F2E) Program, a presentation to Bay Razbani, O., N. Mirzamohammad, and M. Assadi. 2011.
Area Water/Wastewater Workforce, November 14. http:// Literature review and road map for using biogas in internal
baywork.org/wp-content/uploads/2013/12/Central-Marin- combustion engines. Third International Conference on
Food-to-Waste.pdf Applied Energy, May 16–18, Perugia, Italy, http://www.
Drake, N. 2011. CA regulatory arena for dairy based anaero- cense.no/getfile.php/CenSE/Applied%20energy%20paper%
bic digesters. US EPA 6th Annual AgStar Conference, 2001.pdf
Boise, ID, May 10–12. http://www.epa.gov/agstar/docu SCAQMD. 2012. Rule 1110.2—Emissions from gaseous- and
ments/conf11/drake.pdf liquid-fueled engines, amended September 7, 2012. South
EBMUD. 2008. Anaerobic digestion of food waste. Final report Coast Air Quality Management District, Diamond Bar,
prepared by East Bay Municipal Utility District for Funding CA. http://www.aqmd.gov/docs/default-source/rule-book/
Opportunity No. EPA-R9-WST-06-004, March. http://www. reg-xi/rule-1110-2.pdf?sfvrsn=4
epa.gov/region9/organics/ad/EBMUDFinalReport.pdf U.S. Department of Agriculture/Environmental Protection
ECOM. 2016. Technical specifications of ECOM J2KN. Agency/Department of Energy. 2015. Biogas opportu-
ECOM America, Ltd., Gainesville, GA. http://www.sec.nv. nities roadmap progress report. December. http://www.
gov/appeal_docs/refuse_ex7.pdf. rd.usda.gov/files/Biogas-Roadmap-Progress-Report-v12.
Huertas, J.I., N. Giraldo, and S. Izquierdo. 2011 Removal of pdf
H2S and CO2 from biogas by amine absorption. In Mass U.S. Environmental Protection Agency. 2014a. Municipal
transfer in chemical engineering, J. Markos (Ed.). Intech. solid waste generation, recycling, and disposal in the
http://cdn.intechopen.com/pdfs/22869/InTech-Removal_ United States: Facts and figures for 2012. EPA-530-F-14-
of_h2s_and_co2_from_biogas_by_amine_absorption.pdf. 001, February. http://www.epa.gov/solidwaste/nonhaz/
Institute for Local Self-Reliance. 2012. Update of Anaerobic municipal/pubs/2012_msw_fs.pdf
Digester Projects Using Food Waste in North America. U.S. Environmental Protection Agency. 2014b. The benefits
Report prepared by the Institute for Local Self-Reliance of anaerobic digestion of food waste at wastewater treat-
for City of Atlanta, Georgia, October. http://www.ilsr.org/ ment facilities, USEPA Region 9. https://www.epa.gov/
wp-content/uploads/2012/03/atlanta-adreport.pdf sites/production/files/documents/Why-Anaerobic-
Kennedy/Jenks Consultants. 2008. Methane capture feasibility Digestion.pdf
study. Report prepared for City of San Rafael and Central U.S. Environmental Protection Agency. 2014c. Food waste to
Marin Sanitation Agency, December. http://www.cmsa.us/ energy: How six water resource recovery facilities are
assets/documents/administrative/CMSAMethCapture boosting biogas production and the bottom line. EPA
FeasibilityStudyFinalUPDATED010809.pdf 600/R-14/240. https://nepis.epa.gov/Adobe/PDF/
Kraemer, T. 2012. Juicing the low-hanging fruit: Anaerobic diges- P100LDEL.pdf
tion and energy production from MSW Organics in North Warner, D. 2009. Permitting issues for anaerobic digesters in
America. Presentation at SWANA Evergreen Chapter, May the San Joaquin Valley. California Energy Commission
18. http://www.swananw.org/documents/SWANA% Workshop on Biopower in California, April 21. http://
20Presentations/2012-05-18_02-Kraemer_Presentation.pdf www.energy.ca.gov/2009_energypolicy/documents/2009-
Kulkarni, P. 2009. Combined heat and power potential at the 04-21_workshop/presentations/06-San_Joaquin_Air_
California wastewater treatment plants. California Energy District_Presentation.pdf
Commission, CEC-200-2009-014-SF. http://www.energy. Zhang, R., H.M. El-Mashad, K. Hartman, F. Wang, G. Liu, C.
ca.gov/2009publications/CEC-200-2009-014/CEC-200- Choate, and P. Gamble. 2007. Characterization of Food
2009-014-SF.PDF Waste as Feedstock for Anaerobic Digestion. Bioresource
Kuo, J. 2015. Air quality issues related to using biogas from Technol. 98:929–35. http://www.zjubiolab.zju.edu.cn/
anaerobic digestion of food waste. California Energy wumin/userfiles/lab-paper/000292-20101226120531.pdf