ww.sciencedirect.com

b i om a s s a n d b i o e n e r g y x x x ( 2 0 1 4 ) 1e1 1

Available online at w

ScienceDirect

http: / /www.elsevier .com/locate/biombioe

Biomass combustion for greenhouse carbondioxide enrichment

Yves Roy a,*, Mark Lefsrud a, Valerie Orsat a, Francis Filion a,Julien Bouchard a, Quoc Nguyen a, Louis-Martin Dion b, Antony Glover a,Edris Madadian a, Camilo Perez Lee a

aDepartment of Bioresource Engineering, McGill University, Macdonald Campus, Macdonald-Stewart Building,

21,111 Lakeshore Road, Ste. Anne de Bellevue, QC H9X 3V9, Canadab Jean Gobeil & Associes Inc., 1109, de Bromont, Longueuil, QC J4M 2P5, Canada

a r t i c l e i n f o

Article history:

Received 4 October 2013

Received in revised form

25 February 2014

Accepted 1 March 2014

Available online xxx

Keywords:

CO2 enrichment

Biomass

Greenhouse

Furnace

Purification system

Catalytic converter

* Corresponding author. Tel.: þ1 514 441 749E-mail address: yves.roy@mail.mcgill.ca

Please cite this article in press as: Roy Y,Bioenergy (2014), http://dx.doi.org/10.101

http://dx.doi.org/10.1016/j.biombioe.2014.03.0961-9534/ª 2014 Elsevier Ltd. All rights rese

a b s t r a c t

Greenhouses in northern climates have a significant heat requirement that is mainly

supplied by non-renewable fuels such as heating oil and natural gas. This project’s goal

was the development of an improved biomass furnace able to recover the heat and the CO2

available in the flue gas and use them in the greenhouse. A flue gas purification systemwas

designed, constructed and installed on the chimney of a wood pellet furnace (SBI Caddy

Alterna). The purification system consists of a rigid box air filter (MERV rating 14, 0.3 mm

pores) followed by two sets of heating elements and a catalytic converter. The air filter

removes the particulates present in the flue gas while the heating elements and catalysers

transform the noxious gases into less harmful gases. Gas analysis was sampled at different

locations in the system using a TESTO 335 flue gas analyzer. The purification system re-

duces CO concentrations from 1100 cm3 m�3 to less than 1 cm3 m�3 NOx from 70 to

5.5 cm3 m�3 SO2 from 19 cm3 m�3 to less than 1 cm3 m�3 and trapped particulates down to

0.3 mm with an efficiency greater than 95%. These results are satisfactory since they ensure

human and plant safety after dilution into the ambient air of the greenhouse. The recu-

peration of the flue gas has several obvious benefits since it increases the heat usability per

unit biomass and it greatly improves the CO2 recovery of biomass heating systems for the

benefit of greenhouse grown plants.

ª 2014 Elsevier Ltd. All rights reserved.

1. Introduction

The benefits of greenhouse crop production is widely docu-

mented and recognised as a valuable technique to produce

food throughout the year. In addition to the control of the

ambient temperature, a greenhouse allows the producer to

control a variety of other parameters influencing plant growth.

4.(Y. Roy).

et al., Biomass combust6/j.biombioe.2014.03.001

001rved.

The greenhouse can control humidity, temperature, air circu-

lation, water and nutrient cycles, lighting intensity and dura-

tion, and the concentration of carbon dioxide in the ambient

air [1]. The effects of carbon dioxide concentration inside the

greenhouse have been extensively studied in the last two de-

cades in order to determine the optimum level for improved

growth of different plant species [2]. Table 1 summarizes the

benefit of CO2 enrichment for different species of plant.

ion for greenhouse carbon dioxide enrichment, Biomass and

Table 1 e Review of the effect of CO2 enrichment on thegrowth enhancement of different plant species [2].

Observations Growth enhancement

All herbaceous plants þ45%

Woody plants þ48%

Herbaceous populations þ29%

Woody populations þ35%

Dry Matter productions þ20%

Grassland biomass þ12%

Forest Growth þ23%

b i om a s s a n d b i o e n e r g y x x x ( 2 0 1 4 ) 1e1 12

In general, greenhouse enrichment to CO2 concentrations

between 700 and 900 cm3 m�3 is valuable regardless of the

plant species [3]. Some species experience growth reduction

and leaf injuries for concentration higher than 1000 cm3 m�3

[3]. Humans have higher tolerance to carbon dioxide where no

adverse effects are observed for concentrations lower than

7000 cm3 m�3 [4]. As a safety precaution, the Canadian Min-

istry of Health recommends that individuals should not be

exposed to a concentration greater than 3500 cm3 m�3 [4].

The current popularity of biomass furnaces for greenhouse

heating allows for a new source of CO2 for greenhouse pro-

ducers [5]. The utilisation of the CO2 produced during com-

bustion of biomass for greenhouse enrichment has great

advantages for producers since it reduces producers’ environ-

mental footprint and their dependencyon fossil fuelswhichare

typicallyused for enrichment [5].However, theutilizationof the

CO2 from the flue gas of a wood pellet furnace presents

considerable challenges due to the presence of different

noxious gases (CO, NOx, SOx, C2H4, etc) and particulates. Thus,

the aim of this research project was to improve a biomass

furnace in order to recover the heat and theCO2 that is typically

vented with the flue gas and release them into the greenhouse.

2. Flue gas composition

Flue gas composition varies greatly in function of the type of

biomass, thedimensionsof thepellet, puckorbriquette,andthe

typeof stove/furnaceused for thecombustion [6]. Theoxygen to

fuel ratio determines the emission composition during com-

bustion. The expected emissions of different types of stoves/

furnacesaresummarized inTable2 [6]. Ingeneral, pellets stoves

produce less noxious emissions than other stove types because

the shape of the biomass allows a good distribution of oxygen

which produces a uniform and complete combustion. Conse-

quently, pellets stoves flue gas is easier to purify and therefore

more suitable for greenhouseCO2 enrichment. The combustion

Table 2 e Arithmetic average emission levels in mg mL3 at 13%

Appliances Load [kW] Excess airratio

CO [mg/m3]a

[m

Wood-Stoves 9.33 2.43 4986

Fireplace inserts 14.07 2.87 3326

Heat-storing stoves 13.31 2.53 2756

Pellet stove 8.97 3.00 313

a The term m�3 designates a volume at standard reference condition; pr

Please cite this article in press as: Roy Y, et al., Biomass combustBioenergy (2014), http://dx.doi.org/10.1016/j.biombioe.2014.03.001

of all types of biomass results in the production of air-borne

pollutants. The most important noxious gases found in the

flue gas are CO, NOx, particulates, SOx, and VOCs [5]. It is also

common to find trace amounts of nitrous oxides (N2O),

hydrogen chloride (HCl), heavy metals (Cu, Pb, Cd, Hg), poly-

cyclic aromatic hydrocarbons (PAH), polychlorinated dioxins

and furans (PCDD/F), ammonia (NH3) and ozone (O3) [6].

2.1. Carbon monoxide (CO)

Carbon monoxide (CO) is the product of incomplete combus-

tion of carbonaceous fuel into CO2. Thus, CO concentration in

the exhaust gases is a good indicator of the quality of the

combustion [6]. Control of the CO formation can be achieved

mainly through the adjustment of the air to fuel ratio at a

stoichiometric level in the combustion chamber. In fuel lean

conditions, more CO will be formed due to improper mixing

conditions and lack of oxygen to oxidize CO into CO2. In fuel

rich conditions,more COwill be formed in consequence of the

reduction of temperature in the combustion chamber caused

by the excess oxidizer reactant [7]. Increasing the residence

time will also contribute to reduce the concentration of CO in

the emission [8].

It has been reported that a CO concentration of

30e50 cm3 m�3 can be detrimental to plants and cause leaf

chlorosis and abscission as well as flower drop [3]. Also, CO

can negatively affect human health at concentrations lower

than those previously mentioned. Indeed, inhaled CO mole-

cules diminish blood ability to carry oxygen to cell and tissues

[5]. Thus, an exposure of no more than 10 cm3 m�3 of CO

during 24 h is recommended [4].

2.2. Nitrogen oxides (NOx)

Nitrogen oxides are produced by three main gas phase reac-

tion mechanisms during the combustion process. First of all,

there is the fuel NOx mechanism. The nitrogen present in the

wood is volatilized during the combustion process to create

NH3 and HCN molecules [9]. Depending on the fuel type, the

oxygen level, the temperature in the combustion chamber and

the residence time, different NH3 to HCN ratio will be pro-

duced. These molecules are subsequently converted into NO

and NO2 according to a series of elementary reactions [9].

Nitrogen oxides are produced by the thermal NOxmechanism.

At temperatures higher than 1300 �C, nitrogen in the air reacts

with oxygen molecules creating NO [6]. An increase in NO

concentration in the flue gas is generally the result of an

augmentation of temperature, oxygen concentration and

residence time in the combustion chamber. Lastly, NOx are

O2 from small-scale biomass combustion applications [6].

CxHy

g/m3]aParticles[mg/m3]a

NOx

[mg/m3]aTemp [�C] Efficiency [%]

581 130 118 307 70

373 50 118 283 74

264 54 147 224 78

8 32 104 132 83

essure 101.3 kPa and temperature 273 K.

ion for greenhouse carbon dioxide enrichment, Biomass and

b i om a s s a n d b i o e n e r g y x x x ( 2 0 1 4 ) 1e1 1 3

also formed by the prompt NOx mechanism. In fuel-rich

conditions, the nitrogen present in the air can react with CH

radicals produced during the combustion and produce mainly

HCH. These molecules follow the same path as the fuel NOx

mechanism and are converted into NO and N2O [6].

According to these three NOx formation mechanisms, the

reduction of NOx emission can be achieved through a tight

control of the temperature, residence time and oxygen level.

Indeed, keeping the temperature in the combusting chamber

below 1300 �C prevents the formation of thermal NOx. Since

fuel NOx formation is maximized at a temperature of 800 �C, itis beneficial to avoid as much as possible to create a region in

the combustion chamber with such a temperature [9]. Mini-

mizing the residence time in the critical temperature region is

also an important method to control NOx formation. The

stoichiometric ratio has a significant impact on the fuel NOx

reaction. Indeed, at fuel-lean and stoichiometric ratio, NH3

and HCH produce essentially NOwhile in fuel-rich conditions,

the same molecules are converted into N2 molecules [6].

NOx effect on plants varies greatly in function of species.

Mortensen [3] looked at the effect of NOx on 14 species and

discovered that some species are affected by a NOx concen-

tration as low as 0.8e1 cm3 m�3 while other species are not

affected at all. Species that were affected showed leaf chlo-

rosis and necrosis or simply growth-reduction [3]. Moreover,

nitrogen dioxide (NO2) can cause lung damage and negatively

affect the respiratory system of humans at concentration

greater than 0.05 cm3 m�3 (100 mg m�3) [4]. Since nitric oxide

(NO) is usually not found in significant amounts in residential

environment; Canada and the United States have not pub-

lished a maximum limit of exposure for this pollutant. How-

ever, for industrial environments, the US Department of Labor

has determined that the NO concentration in the air should

not exceed 25 cm3 m�3 while exposed during 8 h [10]. It has

been reported that a concentration of 3 cm3 m�3 of nitrogen

oxide has similar effect on the respiratory system than

10e15 cm3 m�3 of CO [5]. Assuming that the NOx concentra-

tion in a greenhouse is composed with 75% NO as estimated

by Mortensen [3], NOx concentration in a greenhouse should

not exceed 4 cm3 m�3 to ensure human safety.

2.3. Sulfur oxides (SOx)

Sulfur oxides are formed during the combustion process by

the complete oxidation of the sulfur contained in the fuel. It

has been reported that 57%e65% of the sulfur present in the

fuel is released into the flue gas while the rest is bound with

the ash [6]. SOx is composed mainly of sulfur dioxide (SO2) at

more than 95%, sulfur trioxide (SO3) at less than 5% and trace

amount is emitted as H2S and K2SO4 [6]. Biomass fuel may

contain as low as 0.01% of sulfur for wood pellets and up to 2%

for other organic feedstocks [5]. Since the amount of SOx

emission is directly correlated to the amount of sulfur in the

fuel, the reduction of fuel sulfur concentration is the primary

method allowing the reduction of SOx emission. Furthermore,

limestone blending with the feedstocks has proved efficient to

reduce SOx flue gas emission by increasing the proportion of

fuel sulfur in the ash [11].

Plants’ SO2 exposure can cause leaf necrosis at a con-

centration greater than 0.5 cm3 m�3 during 4 h [5]. Due to

Please cite this article in press as: Roy Y, et al., Biomass combustBioenergy (2014), http://dx.doi.org/10.1016/j.biombioe.2014.03.001

the considerable solubility of SO2 in water, it can easily

react with the water present in a human’s respiratory track

and irritate during breathing [12]. Thus, an exposure of no

more than 0.019 cm3 m�3 of SO2 during 8 h is recommended

[4,12].

2.4. Particulates

The particulates carried by the flue gas come from different

sources. They can originate from fly-ash, soot, char or

condensed hydrocarbons (tar) [6]. Fly-ash consists primarily of

small ash aggregation carried into the flue gas because of their

low specific density. Fly-ash is also composed of coarse fly-ash

(ash particles with a diameter larger than 1 mm) and aerosol

created during the combustion process by the reaction of K,

Na, Cl or S present in the burning fuel to form salts such as

KCl, NaCl, K2SO4 [6]. Coarse fly-ash traction in the exhaust

gases can be reduced using primary particle emission reduc-

tion measures by optimizing the combustion chamber design

to avoid high turbulence air flow. Aerosol concentration in the

flue gas varies greatly in function of the composition of the

burning fuel. Primary emission reduction measures are not

very effective because of the high volatility of aerosol. Thus

secondary emission reduction measures are more suitable

when aerosol emission is a problem [6].

Particle emissions resulting from incomplete combustion

are composed of soot, char, and tar. Soot particulates

contain mostly carbon and up to 10% hydrogen. Condensed

heavy hydrocarbons (tar) are produced by a local insuffi-

ciency in oxygen in the combustion chamber [13]. Char is

naturally formed in a combustion process during pyrolysis

of the burning fuel. Char’s low specific density makes it

prone to be carried in the flue gas especially at high flue gas

flow rates [6]. Incomplete combustion particles can be

reduced using primary emission reduction measures that

promote complete combustion such as higher combustion

temperature, longer residence time and higher air to fuel

ratio. Additionally, the mixing of steam and carbon dioxide

into the combustion gas can enhance soot oxidation and

therefore reduce soot particle emission [13]. However, sec-

ondary emission reduction devices such as baghouse filters,

cyclone, electrostatic precipitators (ESPs), wet scrubber and

polyethylene membrane separation can be used to reduce

particulate emissions [5,6].

No study on the adverse effect of plant exposure to par-

ticulates has been found in the literature. However, the effect

of particulates on human health has been studied for many

years. Regulatory agencies regulate particulates in function of

their diameter and their concentration in the air. Particulates

that have a diameter larger than 10 mm are typically filtered in

the upper respiratory track [14]. Particulates that have a

diameter between 2.5 and 10 mm can cause nose and throat

irritation, lung damage and bronchitis [12]. It is recommended

to limit the exposure to particle smaller than 10 mm (PM10) at

no more than 150 mg m�3 (24 h) [12]. Particulates that have a

diameter smaller than 2.5 mm (PM2.5) can enter deeply into the

lung. They are strongly associated with cardiovascular and

respiratory mortality and morbidity endpoints [4]. Therefore,

exposure concentration should not exceed 40 mg m�3 (8 h) in

order to ensure human safety [4].

ion for greenhouse carbon dioxide enrichment, Biomass and

Table 3 e Air quality guidelines for a greenhouse environment.

Standards or guideline Humans Plants Humans andplants

Canadian(Health Canada)a

United State(NAAQS/EPA)b

World HealthOrganisation

(WHO)c

Plants maximumexposured

Greenhouseenvironment

Carbon dioxide (CO2) 3500 cm3 m�3 (8 h) e e 1000 cm3 m�3 1000 cm3 m�3

Carbon monoxide (CO) 10 cm3 m�3

(8 h) 25 cm3 m�3 (1 h)

9 cm3 m�3 (8 h)e

35 cm3 m�3 (1 h)e6 cm3 m�3 (24 h)

9 cm3 m�3 (8 h)

31 cm3 m�3 (1 h)

87 cm3 m�3 (15 min)

30 to 50 cm3 m�3 depending

on species

10 cm3 m�3 (8 h)

25 cm3 m�3 (1 h)

NOx Nitrogen oxide (NO) e 3 cm3 m�3 (8 h)f e 0.8e1 cm3 m�3 for some

specific species

0.8 cm3 m�3g

Nitrogen dioxide (NO2) 0.05 cm3 m�3 (8 h)

0.25 cm3 m�3 (1 h)

0.05 cm3 m�3 (1yr) 0.02 cm3 m�3 (1yr)

0.11 cm3 m�3 (1 h)

0.05 cm3 m�3 (8 h)

0.25 cm3 m�3 (1 h)

Sulfur dioxide (SO2) 0.019 cm3 m�3 (8 h)

0.08 cm3 m�3 (5 min)

0.03 cm3 m�3 (1yr)

0.14 cm3 m�3 (24 h)

0.008 cm3 m�3 (24 h)

0.19 cm3 m�3 (10 min)

0.5 cm3 m�3 (4 h) 0.019 cm3 m�3 (8 h)

0.08 cm3 m�3 (5 min)

Particulates (PM2.5) MMADh 40 mg m�3 (8 h)i

100 mg m�3 (1 h)i15 mg m�3 (1yr)

35 mg m�3 (24 h)

10 mg m�3(1yr)

25 mg m�3 (24 h)

e 40 mg m�3 (8 h)

100 mg m�3 (1 h)

Particulates (PM10) MMADh e 150 mg m�3 (24 h) 20 mg m�3 (1yr)

50 mg m�3 (24 h)

e 150 mg m�3 (24 h)

C2H4 e e e 10 mm3 m�3 10 mm3 m�3

a Based on Canadian Residential Indoor Air Quality Guideline [4].b Based on ASHRAE Standard: Ventilation for Acceptable Indoor Air Quality [12].c Based on World Health Organisation [15,17].d Based on two review paper on plant responses to noxious gases [3,5].e Not to be exceeded more than once per year.f This value is not included in the ASHRAE standard. However, under the section covering NOx, it is explained that 3 cm3 m�3 of Nitrogen Oxide has similar effects on the respiratory system than

10e15 cm3 m�3 of CO [16].g This value is obtained assuming a very conservative approach and ismost probably lower than the real safety limit. Since it is not specified inMortensen’s [3] studywhat was the composition of NOx,

it was assumed that it was 100% nitrogen oxide.h MMAD ¼ mass median aerodynamic diameter in microns (micrometers). Less than 3.0 mm is considered breathable; less than 10 mm is considered inhalable.i These standards are based on exposure guidelines for residential indoor air quality (1987) and are only informative. New standard (2012) claims that indoor levels of PM2.5 should be kept as low as

possible. [4,18].

bio

mass

and

bio

energy

xxx

(2014)1e11

4Please

citeth

isarticle

inpress

as:

RoyY,etal.,

Biom

ass

com

bustio

nforgreenhouse

carb

ondioxideenrich

ment,Biom

ass

and

Bioenergy(2014),http

://dx.doi.o

rg/10.101

6/j.b

iom

bioe.2014.03.00

1

b i om a s s a n d b i o e n e r g y x x x ( 2 0 1 4 ) 1e1 1 5

2.5. Volatile organic compounds (VOC)

Volatile organic compounds are any volatile molecule con-

taining organic carbon bound with other carbon, hydrogen,

nitrogen or sulfur [5]. Each type is regulated independently

since their effect on human health varies greatly with their

composition. Every primary emission control measures pro-

moting complete combustion help in reducing VOC’s emis-

sion. Thus VOC should not be a concern if a proper

temperature in the combustion chamber, a long enough

residence time of burning fuel in the combustion chamber and

suitable air to fuel ratio are maintained. Nevertheless,

ethylene, a particular VOC, should be monitored since trace

amounts of this gas into the air can be detrimental to plants

[5]. Ethylene is a plant hormone that stimulates senescence of

the plants, causing leaf chlorosis and abscission as well as

flower drop [3]. Plant response to ethylene varies greatly with

species which makes it difficult to establish a general con-

centration limit. Moreover, studies have shown that the effect

of ethylene concentration on plant varies in function of car-

bon monoxide concentration in the ambient air [15]. Morten-

sen [3] has established that ethylene concentration in a

greenhouse should be kept lower than 10mm3 m�3 in order to

ensure no adverse effects on all species.

3. Air quality guidelines for a greenhouseenvironment

Table 3 summarizes different regulations, guidelines and

research on air quality ensuring safety for human and plants

[3e5,10,12,14,16e18]. Since the present research has been

conducted in Canada, the greenhouse air quality guidelines

are primarily based on Health Canada guidelines and the

maximum exposure concentration for plant safety. As shown

in this table, the World Health Organisation guidelines are

very stringent in terms of air quality while Canadian and US

guidelines are more permissive.

In order to ensure an effortless comparison between the

different standards, each standard in Table 3 has been re-

ported cm3 m�3 for noxious gases and mg m�3 for particulates.

However, some guidelines and research was not published in

those units.

Fig. 1 e Settling chamber [19].

Please cite this article in press as: Roy Y, et al., Biomass combustBioenergy (2014), http://dx.doi.org/10.1016/j.biombioe.2014.03.001

4. Exhaust purification apparatus

The technology to treat or remove noxious gas and particu-

late, can be divided into 5 process classes: mechanical

collection; electrostatic precipitation; filtration, scrubbing,

and catalytic conversion.

4.1. Mechanical collectors

Mechanical collectors are the less expensive type of filtering

devices. However, their efficiency is limited to large scale

particulates [19]. Mechanical collectors can be divided in two

main classes. There are settling chambers (Fig. 1) and cyclone

(Fig. 2). Settling chambers are based on the principle that

particulates will settle due to gravity. The length of the

chamber and the gas velocity are the main factors affecting

the effectiveness of the system. Their main advantages are

that they are simple, inexpensive, cause a small pressure drop

(<20 Pa), have a high capacity and require no water [6]. How-

ever, their efficiency is low and limited to large particles

(�50 mm) and their dimensions limit their utilisation to

constraint environments [19]. Cyclone separators are based on

the principle that the inertia of the particles makes them

collide with the outer wall when a gas stream is spinning in-

side a cylinder. Depending of their design configuration, their

efficiency varies between 50 and 90% and they can remove

particulates of larger diameters than 5e25 mm [19]. Their main

advantages are that they are simple to design and maintain,

they produce low to moderate pressure drop (60e150 Pa), they

can handle high loading and temperature and they are

economical [6]. However, they require head room, they have

low collection efficiency for small particles and they are sen-

sitive to variable dust loading and flow rate.

Fig. 2 e Reverse flow cyclone [6].

ion for greenhouse carbon dioxide enrichment, Biomass and

b i om a s s a n d b i o e n e r g y x x x ( 2 0 1 4 ) 1e1 16

4.2. Electrostatic precipitators

Electrostatic filters operate using the principle that an elec-

trode exerts an attractive force on oppositely charged particles

[19]. Two electrodes highly charged generate ions thatmigrate

fromone electrode to the other. At the same time, the polluted

gas stream flows in between the two electrodes. Thus, dirt

particles collidewith the flowof ions and become charged. The

newlychargedparticles are therefore attracted to theelectrode

due to electrostatic force. Periodically, this electrode is cleaned

using vibration or simply by scraping the electrode [19].

Electrostatic precipitators can achieve up to 99% efficiency

and collect particulates smaller than 1 mm [19]. They have the

ability to collect wet or dry particles, they generate very small

pressure drop (15e30 Pa) and they can operate at high flue gas

flow rates and temperature (up to 480 �C) [6]. Nevertheless,

their initial cost is relatively high, they are sensitive to variable

loading and flow rate and safety measures must be taken to

safeguard personnel from high voltage. Fig. 3 illustrates a cy-

lindrical single-stage electrostatic precipitator.

4.3. Filter

Bag filters operate using the principle that the size of partic-

ulates is greater than that of a gaseous molecule. Their con-

struction is simple and consists of a fibrous or packed media

into which the polluted gas stream is forced to pass through.

Once clogged of particulates, reverse flow can be used to clean

the filtering media and collect the particulates. Most filtering

media need to be replaced occasionally in order to limit

pressure drop and maintain high efficiency. Bag filters can

achieve above 99% efficiency and remove particles smaller

Fig. 3 e Cylindrical single-stage electrostatic precipitator

[19].

Please cite this article in press as: Roy Y, et al., Biomass combustBioenergy (2014), http://dx.doi.org/10.1016/j.biombioe.2014.03.001

than 1 mm in diameter [19]. Bag filters allow high removal ef-

ficiency at low cost. Nonetheless, filtering media can be

damaged or clogged by high temperature, chemical reaction

and humidity and most devices are voluminous [6].

4.4. Scrubber

Scrubbers takeadvantageof the fact thatawaterdropletmoving

in a dirty gas stream collides and intercepts particulates. Thus,

the more droplets sprayed in the gas stream, themore efficient

the scrubber. However, numerous and smaller droplets

generate a larger pressure drop and consume more energy.

Scrubber efficiency varies significantly in function of their

design. In general, spray towers have an efficiency of 80% for

particulates greater than 10 mm, cyclonic scrubbers have an ef-

ficiency of 80% for particulates greater than 2.5 mm, impinge-

mentscrubbershaveanefficiencyof80%forparticulatesgreater

than 2.5 mmand venture scrubbers have an efficiency of 99% for

particulates greater than 0.5 mm [19]. Scrubbers have numerous

advantages since they can simultaneously absorb SO2, NO2, HCl

and particles, they have the ability to cool and clean high-

temperature wet gases and they can treat corrosive gases [6].

However, thenecessity to have awastewater treatment system

to purify the water use in the scrubber is an additional expense

and concern that limit their use. They also have corrosion

problems and they cannot be used in cold weather.

4.5. Catalytic converter

Catalytic converters allow converting noxious gases produced

during the combustion process into less hazardous gases.

Unlike the other purification devices, catalytic converters are

efficient only to process noxious gases and can only work if

the gas stream contains very low particles concentration.

Catalytic converters are madewith a ceramic structure coated

by preciousmetals such as palladium, platinum and rhodium.

Platinum and palladium promote the oxidation of CO and HC,

whereas rhodium promotes the reduction of NOx [20]. In order

to maximize surface area and minimise the quantity of

precious metal used, the catalyst are usually shaped like

honeycombs, either squares or triangles, in order to get more

surface area. A drawback of catalytic converters is that they

are required to operate at a high temperature in order to be

efficient. In fact, the activation energy for Pt catalysed CO to

CO2 is about 83.68 kJ mol�1, compared to 167.36 kJ mol�1 for

non-catalysed or thermal reaction [21]. The catalysed reaction

of CO into CO2 has an activation temperature of approxi-

mately 200 �C. Thus, the temperature must be carefully

controlled within the catalytic converter.

5. Materials and methods

A wood pellet furnace was installed inside a tunnel green-

house and an emission purification system was installed on

the chimney of the biomass furnace. The composition of the

flue gas produced during wood pellet combustion was ana-

lysed and recorded. The specification of the furnace, wood

pellets used, and testing instruments are provided in the

following section.

ion for greenhouse carbon dioxide enrichment, Biomass and

b i om a s s a n d b i o e n e r g y x x x ( 2 0 1 4 ) 1e1 1 7

5.1. Furnace

A biomass furnace (SBI Caddy Alterna, Saint-Augustin-de-

Desmaures, Quebec, Canada) was installed inside a tunnel

greenhouse located on the Macdonald campus of McGill Uni-

versity (Ste-Anne-de-Bellevue, Quebec, Canada). The furnace

is equipped with a 101.6 mm chimney, a 500 Watts lighter, a

central computer allowing to control the input power which

can be set to 4.98, 17.58, 23.45, 29.31 and 35.17 kW [22]. Based

on dealer specifications, this furnace has an average thermal

efficiency of 81.2% [22].

5.2. Wood pellets

The biomass used during the experiment was premium grade

wood pellets made with 100% hardwood. As specified by the

manufacturer (Valfei Product Inc, Quebec, Canada), the wood

pellets contain no additive and chemical and produce less

than 1% of ash and less than 0.5% of fines. The bulk density of

hardwood pelletizedwoodwasmeasured at 580 kgm�3.Wood

pellets have a cylindrical shape with an average diameter of

8 mm and a length of 30 mm.

5.3. Gas analysis

The composition of the flue gas produced during wood pellet

combustionwasanalysedusingtheTESTO335portableanalyzer

(Testo Inc., Lenzkirch, Germany). This analyser has a resolution

of100mm3m�3 forCO,NOandNO2andaresolutionof1cm3m�3

for SO2 [23]. VOC analyses were performed using VOC detectors

(Reed GD-3300, Taipei, Taiwan) which have sensitivity of

50 cm3 m�3 for methane [24]. Volatile organic compound anal-

ysis was required in order to detect the presence of ethylene

which can be detrimental to plant as previously stated.

5.4. Experimental setup

The purpose of this research was to develop an exhaust

filtering system that can be installed on a residential size

wood pellet furnace for small scale greenhouse enrichment.

Fig. 4 e Prototype and designed gener

Please cite this article in press as: Roy Y, et al., Biomass combustBioenergy (2014), http://dx.doi.org/10.1016/j.biombioe.2014.03.001

The purification system effectiveness and cost were themajor

criteria considered during the design of the system. The first

challenge was to remove particulates present in the flue gas.

The particulates filtering system needed to handle an average

flue gas temperature of 215 �C (at the maximum input power

of 35.17 kW). Among the different particulate removal appa-

ratus, the air filtermembranewas themost convenient choice

since it was easy to install and provides high removal effi-

ciency at low cost. A catalytic converter is the most cost-

effective apparatus able to eliminate all major noxious gases

present in the flue gas. Fig. 4 shows the assembled prototype

system in the experimental greenhouse of McGill University

as well as the design drawing.

5.5. Purification system

The purification apparatus included an air filter, catalytic

converters, heating elements and fans. The assembling details

are shown in Fig. 5. The purification system cannot allow any

flue gas leakage outside the system and all exhaust gases

must go through the system (including filtration) before

entering the greenhouse. Thus, all the components were

installed inside a galvanized steel frame seal using metallic

sealing tape along all edges and possible leak points.

The first filtering component of the purification system

consisted of a rigid-box air filter using fibreglass as filtering

media. This high efficiency metal box air filter can trap

contaminant down to 0.3 mm and have efficiency greater than

95% [25].

Following the filter, two sets of heating element and two

catalytic converters were installed in the system to treat any

noxious gases. The heating elements were required to control

the temperature of the flue gases before entering the catalytic

converters. The flue gas temperature at the exit of the furnace

varied greatly in function of the ambient temperature which

required the operator to adjust the heating element to obtain a

constant temperature in front of the catalytic converter.

Consequently, the first heating element and catalytic con-

verter was set to heat the air to a temperature of 230 �C, which

from internal testing appears to be theminimum temperature

al purification system assembly.

ion for greenhouse carbon dioxide enrichment, Biomass and

Fig. 5 e Slice of the purification system.

b i om a s s a n d b i o e n e r g y x x x ( 2 0 1 4 ) 1e1 18

for CO and SO2 removal using a universal type catalyser con-

taining a loading of 20.34 g m�3 of Palladium and no Rhodium

and Platinum (Catalyseur National, St-Hubert, Quebec, Can-

ada). Increasing the air temperature above 250 �C increased

the production of thermal NOx by the first heating element.

The second heating element raised the flue gas temperature to

350 �C in order to improve the efficiency of the second cata-

lytic converter to process NOx into nitrogen and oxygen.

It is essential to properly install the high efficiency air filter

prior to the catalytic converter because particulates reduce

the contact area between noxious gases and the precious

metal coating inside the catalytic converter which signifi-

cantly reduce noxious gas conversion. Our testing has shown

that the usage of a fibreglass air filter pad, with a removal

efficiency of 80%, will cause the catalytic converter to lose half

of its efficiency to remove CO in only 36 h of operation. Thus,

particulates removal efficiency prior to the catalyser must be

monitored and controlled to ensure plants and humans

safety.

Fig. 6 e Gas concentration measured before the purific

Please cite this article in press as: Roy Y, et al., Biomass combustBioenergy (2014), http://dx.doi.org/10.1016/j.biombioe.2014.03.001

The high restriction caused by the catalytic converters and

the air filter made it necessary to install a fan at the entrance

of the flue gas filtering system to push the air through the

filtering system. Due to the high temperature of the flue gas, it

was not possible to use a regular centrifugal fan without

causing heat damage to the motor. In order to protect the

motor of the fan from the heat of the flue gas, a shaft was

machined and installed between the fan propeller and the

motor. This shaft protected the motor from overheating by

blowing air on themachined shaft using a regular ventilator. A

second fan was installed at the exit of the system to enhance

the pressure differential through the system allowing a better

flow of flue gas (Fig. 5).

6. Results and discussion

Figs. 6 and 7 show noxious gases and temperature variation at

the entrance and the exit of the purification system attached

ation system attached on the wood pellet furnace.

ion for greenhouse carbon dioxide enrichment, Biomass and

Fig. 7 e Gas concentration measured after the purification system attached on the wood pellet furnace.

b i om a s s a n d b i o e n e r g y x x x ( 2 0 1 4 ) 1e1 1 9

on the wood pellet furnace. It is important to point out that

during each analysis, the probe was inserted inside the

chimney at time 0 min for CO, NO, NO NO2 and temperature

analysis and at time 30 min for SO2 and NO2. This situation

explains the rapid variation on the left hand side of each

figure. SO2 and NO2 analysis have not been taken at the same

time as the other analysis because of the limitation in the

number of analysis samples that the Testo analyser can

handle simultaneously. Thus, the SO2 and NO2 graphs cannot

be overlapped over the other graphs because of time

discrepancy.

Fig. 6 illustrates the variation of the main noxious gases in

the flue gas. Gas analysis presented in this figure was taken in

the chimney of the furnace after 2 h of operation. It is possible

to conclude that the NOx are mainly composed of NO. NOx

concentration in the flue gas during this trial was on average

66 cm3 m�3 with a standard deviation of 15 cm3 m�3 while NO

and NO2 concentration was on average

63 cm3 m�3 � 15 cm3 m�3 and 0.35 cm3 m�3 � 0.17 cm3 m�3.

SO2 flue gas concentration was on average

9 cm3 m�3 � 2.4 cm3 m�3 which is typical for this kind of fuel.

Also, this figure shows that CO concentration varies greatly

during the combustion operation. During this trial, CO

Table 4 e Emission concentrations in function of the sampling

Gases Flue gas concentration[cm3 m�3]

Exit of pur[c

CO 1100 w0 (at cm3 m

NO 64 5.2

NOx 70 5.5

SO2 19 w0 (at cm3 m

VOC N/A Not detectable

Particles N/A Not visible

a Since CO and SO2 were not detected at ppm level, the calculations wer

Please cite this article in press as: Roy Y, et al., Biomass combustBioenergy (2014), http://dx.doi.org/10.1016/j.biombioe.2014.03.001

concentration in the flue gas was on average

955 cm3 m�3 � 290 cm3 m�3 but reached concentration as high

as 1600 cm3 m�3. This result might be a sign of improper

mixing of air andwood pellets or insufficient residence time in

the reactor of the furnace. However, internal modification of

the furnace was not part of this research project and the focus

was made on the design of a purification system that can

accommodate those kinds of variations. Operational flue gas

temperature was very stable at 177 �C. It was necessary to

increase this flue gas temperature in order to enhance the

efficiency of the catalytic converters. The adjustment of the

flue gas temperature was performed using two electrical

heating elements because they are easy to adjust and they can

be attuned during furnace operation.

Fig. 7 shows noxious gases and temperature variation at the

exit of the purification system attached on the wood pellet

furnace. The gas analysis presented below was taken directly

from theexhaust of the last fan of thepurification system5min

after start-up of the purification system. Similarly to the flue

gas, NOx in the purified flue gas are mainly composed of NO

since NOx concentration during this trial was on average

8.9 cm3 m�3 with a standard deviation of 3.3 cm3 m�3

while NO and NO2 concentration was on average

location.

ification systemm3 m�3]

Ambient air of greenhouse[cm3 m�3]

�3 level) <0.01a

0.1

0.11�3 level) <0.01a

using Reed GD-3300 Not detectable using Reed GD-3300

Not visible

e made assuming a concentration of 0.5 cm3 m�3.

ion for greenhouse carbon dioxide enrichment, Biomass and

b i om a s s a n d b i o e n e r g y x x x ( 2 0 1 4 ) 1e1 110

8.5 cm3m�3� 3.1 cm3m�3 and 3.4 cm3m�3� 0.7 cm3m�3. It has

beenobserved thatNOxconcentrationcanvarydepending if the

analysis is performedduring the first 30e45min of operation or

after. This isbecausecatalytic convertersneed tobehot inorder

to be efficient and 30e45 min is the time required to reach the

lower limit operational temperature. In order to reduce the

warmup time, itwould benecessary tomodify the set-up of the

furnace to increase exhaust flue gas temperature. Another so-

lutionwould be to install a bypass at the exit of the purification

rejecting thefirsthourofgasproduced into theatmosphere. SO2

concentration at the exit of the purification system was on

average 0.02 cm3 m�3 � 0.14 cm3 m�3 which is under the

1 cm3 m�3 resolution for this instrument. CO concentration in

the flue gaswas on average 1.1 cm3m�3� 0.8 cm3m�3 which is

very closeof the resolution limit of 1 cm3m�3 of this instrument

even though gas analysis has been conducted during thewarm

up of the purification unit. The catalytic converter for CO

removal produced a remarkable efficiency reducing CO con-

centration from 955 cm3 m�3 to only 1.1 cm3 m�3.

Seven experiments of approximately 4 h each have been

conducted on the purification system. Averaging all the re-

sults obtained during these multiple trial, it was found that

the emission control system installed on the chimney of the

furnace was able to reduce CO concentrations from 1100 to

less than 1 cm3 m�3, NOx from 70 to approximately

5.2 cm3m�3 and SO2 from 19 to less than 1 cm3m�3. Therewas

no significant variability among the different trial but it was

observed that the first hour of operation released generally

higher concentrations of NOx than after 1 h of operation. The

purification system was able to trap particles down to 0.3 mm

with an efficiency greater than 95%. In addition, volatile

organic compounds were not detected for any of the trials at

the exit of the system using the Reed GD-3300 analyser, which

confirms that no significant amount of ethylene is released

into the greenhouse.

The furnace was installed in a double layer polyethylene

greenhouse that has an approximate volume of 370 m3. This

size of greenhouse will experience approximately 0.5 air

changes per hour, excluding the effect of the ventilation [26].

At maximum input power of 35.17 kW, the volumetric flow

rate of purified flue gas was on average 0.001 m3 s�1. The

naturally occurring dilution factor was calculated using Eq.

(1) and was calculated to be approximately 50.

Dilution factor ¼ Greenhouse air exchange flow ratePurified flue gas flow rate

(1)

Table 4 shows the concentrations of noxious gases at the

entrance and the exit of the purification system as well as

their theoretical concentration in the ambient air of the

greenhouse. The ambient air concentrations of noxious

components are significantly below limit exposure values for

humans and plants reported in Table 3. These results are

satisfactory since they ensure human and plant safety.

7. Conclusion

The current study has shown that a purification system

composed of a rigid box air filter (MERV rating 14, 0.3 mm

Please cite this article in press as: Roy Y, et al., Biomass combustBioenergy (2014), http://dx.doi.org/10.1016/j.biombioe.2014.03.001

pores), two sets of heating element, two catalytic converters,

with two forced air fans is a suitable system to filter wood

pellet furnace exhaust gases for greenhouse CO2 enrich-

ment. Analysis shows that the purification system installed

on the chimney of the furnace was able to reduce CO con-

centration from 1100 to less than 1 cm3 m�3, NOx from 70 to

5.5 cm3 m�3, SO2 from 19 to less than 1 cm3 m�3 and

removed particulates larger than 0.3 mm with an efficiency

greater than 95%. Volatile organic compounds were not

detected for any of the trials at the exit of the system which

confirms that no significant amount of ethylene is released

into the greenhouse. These results meet and exceed the air

quality criterion for greenhouse environments after dilution

into the ambient air of the greenhouse. The designed system

provides an alternative to current carbon dioxide enrich-

ment technologies which uses fossil fuel to provide CO2 for

most greenhouse operations and considerably improved the

thermal efficiency of wood pellet heating system since no

heat is lost trough the flue gas.

Acknowledgements

We would like to thank SBI Company for providing a wood

pellet furnace for this project, Valfei Product for the wood

pellets, and Catalyseur National Company, for their technical

support on catalytic converter. Natural Sciences and Engi-

neering Research Council of Canada, Fonds quebecois de la

recherche sur la nature et les technologies, BioFuelNet (29)

and Ministere de l’Agriculture, des Pecheries et de l’Alimen-

tation duQuebec (MAPAQ) (809142) are acknowledged for their

financial support. Finally, I would like to give my sincere ac-

knowledgements to the Bioresource Engineering Department

staff andmore specifically to ScottManktelow andDr. Samson

Sotocinal for their technical help during the construction of

the system.

r e f e r e n c e s

[1] Mortensen LM. Growth responses of some greenhouse plantsto environment. I. Experimental techniques. Sci Hortic1982;16(1):39e46.

[2] Kirschbaum MU. Does enhanced photosynthesis enhancegrowth? Lessons learned from CO2 enrichment studies. PlantPhysiol 2011;155(1):117e24.

[3] Mortensen LM. Review: CO2 enrichment in greenhouses.Crop responses. Sci Hortic 1987;33(1e2):1e25.

[4] Health-Canada. Residential Indoor Air Quality Guidelines[Guideline on the internet]. Ottawa: Health Canada; 2012.Available from:, http://www.hc-sc.gc.ca/ewh-semt/air/in/res-in/index-eng.php [cited 27.01.14].

[5] Dion LM, Lefsrud M, Orsat V. Review of CO2 recoverymethods from the exhaust gas of biomass heating systemsfor safe enrichment in greenhouses. Biomass Bioenergy2011;35(8):3422e32.

[6] Van Loo S, Koppejan J. The handbook of biomass combustionand co-firing. London, UK: Earthscan; 2008. p. 465.

[7] Roy MM, Dutta A, Corscadden K. An experimental study ofcombustion and emissions of biomass pellets in a prototypepellet furnace. Appl Energy 2013;108(0):298e307.

ion for greenhouse carbon dioxide enrichment, Biomass and

b i om a s s a n d b i o e n e r g y x x x ( 2 0 1 4 ) 1e1 1 11

[8] Godbout SH, Palacios J, Larouche J-P, Brassard P, Pelletier F.Bilan energetique, emissions gazeuses et particulaires de lacombustion de la biomasse agricole a la ferme. Quebec,Canada: Institut de recherche et de developpement enagroalimantaire (IRDA); 2012. p. 239 [Project number IRDA#:100 035].

[9] Winter F, Wartha C, Hofbauer H. NO and N2O formationduring the combustion of wood, straw, malt waste and peat.Bioresour Technol 1999;70(1):39e49.

[10] OSHA (Occupational Safety and Health Standards). Table Z-1Limits for air contaminants [Guideline on the internet].Washington, DC: Occupational Safety and HealthAdministration (OSHA); 2006. Available from:, www.osha.gov/pls/oshaweb/owadisp.show_document?p_table¼STANDARDS&;p_id¼9992 [cited 27.01.14].

[11] Lupianez C, Guedea I, Bolea I, Dıez LI, Romeo LM.Experimental study of SO2 and NOx emissions in fluidizedbed oxy-fuel combustion. Fuel Process Technol2013;106(0):587e94.

[12] ASHRAE. ASHRAE Standard: ventilation for acceptableindoor air quality. ANSI/ASHRAE Addendum q to ANSI/ASHRAE Standard 621. Atlanta, GA, United States: AmericanSociety of Heating, Refrigerating and Air-ConditioningEngineers; 2010. p. 18.

[13] Law CK. Combustion physics. Cambridge, NY, United States:Cambridge University Press; 2006. p. 722.

[14] ASHRAE. 2009 ASHRAE handbook fundamentals. Atlanta,GA, United States: American Society of Heating, Refrigeratingand Air-Conditioning Engineers (ASHRAE); 2009. p. 926.

[15] Hanan JJ. Greenhouses: Advanced Technology for ProtectedHorticulture. Boca Raton: CRC Press; 1998.

[16] WHO. Air Quality guidelines, Global update 2005: particulatematter, ozone, nitrogen dioxide and sulfur dioxide.

Please cite this article in press as: Roy Y, et al., Biomass combustBioenergy (2014), http://dx.doi.org/10.1016/j.biombioe.2014.03.001

Copenhagen, Denmark: World Health Organization (WHO);2006. p. 496.

[17] WHO. WHO guidelines for indoor air quality: selectedpollutants. Copenhagen, Denmark: World HealthOrganization (WHO); 2010. p. 484.

[18] Health-Canada. Exposure guidelines for residential indoorAirQuality: a Report of the Federal-Provincial AdvisoryCommittee on Environmental and Occupational Health.Ottawa, Canada: Health Canada; 1987. p. 28. Report no.: H46-2/90-156E.

[19] Flagan R, Seinfeld JH. Fundamentals of air pollutionengineering. New Jersey, US: Prentice Hall; 1988. p. 554.

[20] Sideris M. Methods for monitoring and diagnosing theefficiency of catalytic converters a patent-oriented survey.New York, US: Elsevier; 1998.

[21] Heck RM, Farrauto RJ, Gulati ST. Catalytic air pollutioncontrol: commercial technology. New York: Wiley-Interscience; 2002.

[22] PSG. Instruction pour l’installation et le fonctionnement dela fournaise a granule CADDY ALTERNA: Manueld’instruction. Saint-Augustin de Desmaures, Canada: PSG;2011. p. 54.

[23] Testo. Testo 335 combustion analyser. Flanders, NJ, US:Testo.inc; 2011.

[24] Reedinstrument. GD-3300 Combustible Gas Detector. Ste-Anne-de-Bellevue, Canada: Reed instrument.inc; 2013.

[25] McMaster-Carr. High-efficiency metal box air filter, Merv 14,Aluminum-coated steel, 12 � 24, 12" thick [Specification onthe internet]. Aurora, OH, US: McMaster-Carr.inc; 2013.Available from:, http://www.mcmaster.com/#8231t42/¼qt0f2u [cited 27.01.14].

[26] Hicklenton PR. CO2 enrichment in the greenhouse: principlesand practice. Portland, Or, US: Timber Press; 1988.

ion for greenhouse carbon dioxide enrichment, Biomass and