DOI: 10.3303/CET2292067 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
Paper Received: 15 February 2022; Revised: 11 March 2022; Accepted: 30 April 2022 
Please cite this article as: Caposciutti G., Antonelli M., Barontini F., Galletti C., Tognotti L., Desideri U., 2022, An Experimental Investigation 
on the Effect of Exhaust Gas Recirculation in a Small-scale Fixed Bed Biomass Boiler, Chemical Engineering Transactions, 92, 397-402  
DOI:10.3303/CET2292067 
  

 CHEMICAL ENGINEERING TRANSACTIONS  
 

VOL. 92, 2022 

A publication of 

 

The Italian Association 
of Chemical Engineering 
Online at www.cetjournal.it 

Guest Editors: Rubens Maciel Filho, Eliseo Ranzi, Leonardo Tognotti 

Copyright © 2022, AIDIC Servizi S.r.l. 

ISBN 978-88-95608-90-7; ISSN 2283-9216 

An Experimental Investigation on the Effect of Exhaust Gas 

Recirculation in a Small-Scale Fixed Bed Biomass Boiler 

Gianluca Caposciuttia, Marco Antonellia, Federica Barontinib,*, Chiara Gallettib, 

Leonardo Tognottib, Umberto Desideria 

a Dipartimento di Ingegneria dell'Energia, dei Sistemi, del Territorio e delle Costruzioni, Università di Pisa, Largo Lucio 

Lazzarino 2, 56122 Pisa, Italy  
b Dipartimento di Ingegneria Civile e Industriale, Università di Pisa, Largo Lucio Lazzarino 2, 56122 Pisa, Italy 

 federica.barontini@unipi.it 

Exhaust gas recirculation is a technique that allows for controlling the combustion chamber temperature and 

reducing the NOx and particle matter emissions. Moreover, it helps to mitigate soot formation and ash 

agglomeration in combustion systems. The present study investigated the effect of exhaust gas recirculation on 

combustion temperatures of a 140 kW underfed stoker biomass boiler. To this purpose, a wide range of 

operating conditions were used, collecting data regarding flue gas and fixed bed temperatures. It turned out that 

the recirculating ratio has a significant effect on the temperatures in the primary combustion zone, affecting the 

thermal gradient and the main thermal zones of the biomass combusting bed. The obtained results can be useful 

for lumped parameter modeling, or CFD validation purposes. 

1. Introduction 

With the increment in the world energy demand and the increasing concerns on environmental pollution, global 

warming, and climate change, the use of renewable, sustainable, and environmentally-friendly energy resources 

has become mandatory. In this scenario, biomass appears as one of the most promising renewable energy 

resources, because of its abundance, wide distribution, and carbon dioxide neutrality. Among all biomass-to-

energy conversion technologies, biomass combustion for heat and power generation represents the most 

common and efficient conversion route. Stoves and boilers fed with biomass woodchips have a wide broad 

market share in Europe, especially for small-power applications. However, further efforts in the design and 

operation of these devices are still required to improve their performance in terms of efficiency and pollutant 

emissions. In particular, the NOx, CO and particulate matter emissions pose a growing concern for the recent 

European regulations in terms of air quality. Therefore, several strategies have been proposed to reduce 

pollutant emissions from such devices, such as air staging and exhaust gas recirculation. Amongst them, 

exhaust gas recirculation (EGR) can provide key advantages. Firstly, relative air-fuel velocity and combustion 

turbulence can be increased without increasing the oxygen excess in the combustion chamber. Secondly, the 

fuel bed temperature can be limited thanks to the reduced oxygen availability, thus reducing NOx production, 

also mitigating ash agglomeration and soot transport phenomena.  

However, the potential of the EGR technique has been scarcely explored in the literature. Schulte et al. (2020) 

proved that the use of EGR in combination with a low excess air ratio in the fuel bed, could reduce particulate 

matter emissions from a pellet boiler. Archan et al. (2021) developed and experimentally investigated a small-

scale (200 kW) multi-fuel biomass grate furnace, employing a low oxygen concentration in the bed and a double 

air staging, including the supply of flue gas recirculation. Their results indicate a positive influence of recirculated 

flue gas on NOx emissions, since the EGR utilization within the reduction zone limits the NOx formation potential 

by minimizing local temperature peaks and lowering the oxygen partial pressure. Numerical models, specifically 

Computational Fluid Dynamics (CFD) models, were applied to the theoretical study of the effect of EGR on 

boiler performance and gas contaminant emissions. Gomez et al. (2019) carried out several simulations of a 30 

kW boiler operating in different conditions: the results showed that EGR can increase the boiler thermal 

397



performance and reduce the NOx emissions, especially for low oxygen excess values. The results of the CFD 

modelling study of a 13 MW waste wood-fired grate boiler performed by Rajh et al. (2018) indicated that an 

optimized configuration of the air and recycled flue gas jets can greatly enhance mixing and extend the residence 

time of the combustibles in the hot zones in the furnace, and therefore improve combustion and reduce pollutant 

emissions, especially those due to incomplete combustion such as CO. The availability of specific experimental 

data is crucial for the development and validation of numerical models, that could be used to improve the 

performance of biomass boilers (Patronelli et al., 2018). 

In the present work, the effect of the EGR was experimentally studied in a 140 kW fixed-bed biomass boiler. To 

this purpose, the boiler was equipped with control and measuring systems to analyze temperature, mass flow 

and gas composition. More specifically, the temperatures in the fuel bed and the exhaust gas composition were 

monitored under different operating conditions, i.e., equivalence ratio, primary to secondary air-flow ratio, and 

EGR ratio, to assess the potential of EGR in biomass furnaces. 

2. Method 

The fixed bed boiler is a 140 kW underfed stocker belonging to the University of Pisa and is shown in Figure 1a. 

The fixed bed has a 600 mm length, 150 mm width, and 200 mm depth, and it is equipped with a screw feeder 

(having an 80 mm diameter and 80 mm step) providing the biomass flow 𝑚𝐵 at a constant rate from below the 

bed (see Figure 1b). The total air flow is supplied by a 2.2 kW centrifugal blower, and it is split into primary and 

secondary flows using two sphere-valves. The primary air is injected 20 mm below the biomass fixed bed surface 

through 68 rectangular nozzles (20 mm length and 3 mm height), distributed in 3 rows around the fixed bed. 

The secondary air is fed around the biomass bed through a stainless-steel pipe system and injected by 7 circular 

ducts (20 mm diameter). An exhaust gas blower extracts the flue gases from the combustion chamber and to 

keep the pressure at 20 Pa below the ambient one for safety reasons. After the combustion stage, a gas-to-oil 

heat exchanger heats up a Seriola 1510 diathermic oil, thus transferring the heat to a Kettle water boiler. 

More details about the boiler are given in previous works (Caposciutti et al., 2018, 2020); however, with respect 

to these works the boiler was modified here to recirculate the exhaust gases. More specifically, the exhaust 

gases are extracted downstream of the heat exchanger, with a 𝑚𝑅  mass flow rate, and subsequently are mixed 

with the primary air. The value of 𝑚𝑅 is controlled by means of a 0.75 kW high-temperature blower and a gate 

valve, while 𝑚𝑅  is evaluated through a stagnation pressure measurement system. 

The primary and secondary air mass flow rates, i.e., 𝑚𝑎,𝑝 and 𝑚𝑎,𝑠, respectively, are measured by using two 

TFLOW T112 hot wire sensors. The heat exchanger gas inlet and outlet temperatures (namely 𝑇𝑓,𝑖𝑛 and 𝑇𝑓,𝑜𝑢𝑡 , 

respectively) are monitored by means of radiation-shielded k-type thermocouples. The flue gas composition, in 

terms of CO2 and O2, is monitored by means of a NDIR and a paramagnetic-based instruments from 

Environnement SA for mass balance calculation. A scheme of the measuring system is provided in Figure 2a. 

The fixed bed is instrumented with 12 k-type thermocouples to measure the temperature in several positions, 

as described by the scheme in Figure 2b. These are mounted on a stainless-steel housing to keep them in 

position during the operation. The thermal sensor housing is designed to limit the interaction with the biomass 

and gas flow. 

 

(a)   (b)   

Figure 1: Picture of the boiler (a) and fixed bed (b). 

 

398



 
(a) 

 
(b) 

Figure 2. Scheme of the system (a) and of thermocouples displacement (i.e., black dots) in the fixed bed (b). 

For the thermocouple displacement, three ZY planes were considered with different distances from the biomass 

flow inlet, thus the Left, Central, and Right planes were chosen. The distance of the Left plane from the biomass 

inlet was 315 mm. Three lines on the ZX fixed bed symmetry plane were chosen to be 100 mm below the 

biomass surface (i.e., lower line), on the biomass surface (i.e., middle line), and 150 mm above it (i.e., upper 

line), the latter corresponding to the secondary air inlet position. Finally, a side-line was defined on the biomass 

surface, i.e., in the XY plane, at a distance of 55 mm from the middle line.  

Hence, the fixed bed thermal behavior was reconstructed by using linear interpolation of the temperatures 

measured in the positions shown in Figure 2b, allowing to evaluate this behaviour for different operating 

conditions. 

The air excess ratio is  

𝜆 =  
𝛼

𝛼𝑠𝑡
   (1) 

where 𝛼 represents the air-to-fuel ratio  

𝛼 =
𝑚𝑎,𝑝 + 𝑚𝑎,𝑠

𝑚𝑏
 (2) 

In the present system, 𝛼 can be directly derived from the flue gas oxygen content, and 𝑚𝑏 is calculated from the 

air-to-fuel ratio knowing the total air flow rate (Caposciutti et al., 2018). 

Moreover, the primary air excess ratio can be defined as : 

𝜆𝑝 =
𝑚𝑎,𝑝

𝑚𝑏
 

1

𝛼𝑠𝑡
=  

𝛼𝑝

𝛼𝑠𝑡
 (3) 

Finally, the recirculation ratio R represents the ratio between the recirculated and total flue gas mass flow, i.e.: 

𝑅 =  
𝑚𝑅
𝑚𝐹

 (4) 

 

The biomass was poplar woodchips whose composition is reported in Table 1. 

399



Table 1: Biomass characteristics. 

C H N O Moisture Ash LHV 

(%daf) (%daf) (%daf) (%daf) (%ar) (%db) (MJ/kg) 

48.76 5.68 0.21 45.35 10.06 0.9 17.88 

 

𝜆 , 𝜆𝑝, and R are the main operation parameters. Tests were performed by varying both 𝜆 and 𝜆𝑝. To evaluate 

the effect of R, all the tests were repeated at the approximately same (𝜆, 𝜆𝑝) conditions, while the gate valve 

was controlled to ensure R to be around 0, 0.3 and 0.6. Totally, 12 operating points were investigated. All the 

temperature and mass flow measurements were managed by means of a NI9214, and a NI9207 boards with a 

NI9223 chassis, and a Labview-based software. Data were sampled for at least 30 min in steady state condition. 

3. Results and discussion 

The resulting operating conditions and flue gas temperatures measured in the experimental tests are reported 

in Table 2, where the different IDs refer to different operating conditions. Consistently with the literature (Carroll 

et al., 2015; Lamberg et al., 2011), 𝜆 was varied between 1.1 and 3, while 𝜆𝑝 between 0.1 and 0.6. To assess 

the effect of R, the tests were replicated at the approximately same (𝜆, 𝜆𝑝) conditions, with different R values, 

namely around 0, 0.3 and 0.6. 

Table 2: Obtained operating conditions 

 ID1 ID2 ID3 ID4 ID5 ID6 ID7 ID8 ID9 ID10 ID11 ID12 

𝜆 (-) 1.21 1.44 1.24 1.35 1.41 1.96 1.45 1.10 1.20 2.53 1.93 3.27 

𝜆𝑝 (-) 0.50 0.43 0.28 0.15 0.35 0.25 0.13 0.39 0.56 0.31 0.23 0.20 

𝑅 (-) 0 0 0.46 0.55 0 0.40 0.48 0.57 0.76 0 0.34 0.30 

𝑇𝑓,𝑖𝑛 (°C) 669 696 698 728 766 722 763 838 830 698 744 686 

 

Figure 3 shows the average fuel bed temperature as a function of 𝜆𝑝, and R. The average fuel bed temperature 

is obtained by averaging the 9 temperature measurements in the bed.  It can be observed how such temperature 

increases with the recirculation ratio, which mixes a high temperature stream with the primary air feeding the 

bed. For low R values, results indicate a non-monotic trend of the average temperature with  𝜆𝑝 thus indicating 

the presence of an optimal value of such parameter to trigger combustion. For low 𝜆𝑝 and R, combustion is 

hampered by the low mixing and oxygen availability; further increasing  𝜆𝑝 the average temperature increases 

up to a maximum value. Subsequently, the average temperature decreases because of the dilution with primary 

air.  With increasing R such behaviour is hampered by the increased mixing and turbulence in the reaction 

region which promote combustion and lead to high temperatures. 

 

 

Figure 3. Average fuel bed temperature vs 𝜆𝑝 and R. 

400



Figure 4 shows the temperatures in the biomass primary combustion area. The top panels refer to the vertical 

ZX plane, while the bottom panels to the horizontal YX plane (see Figure 2b). Six conditions in terms of 𝜆, 𝜆𝑝 and 

R are selected, even though two pairs conditions match each other closely and are devised to assess the 

accuracy of the measurements. 

 

  
(a) (b) 

  
(c) (d) 

  
(e) (f) 

Figure 4. Temperature spatial distribution in the biomass primary combustion area: ZX (top panel) and YX 

(bottom panel) planes. a) ID1; b) ID2; c) ID4; d) ID7; e) ID6; f) ID9. See Table 2 for more details on operating 

conditions. 

 

Indeed, Figure 4 shows good repeatability in temperature distributions obtained with similar operating 

conditions, as shown by comparing Figure 4a (ID1) with Figure 4b (ID2) and Figure 4c (ID4) with Figure 4d 

(ID7). The temperature distribution showed a larger disuniformity when no recirculation was used. When R was 

401



increased, the temperature on the bed surface was more uniform and its average temperature reached its 

maximum. This fact was verified at constant air excess ratio, both by increasing R and lowering 𝜆𝑝 (Figure 4a-

b against Figure 4c-d, i.e. ID1-ID2 against ID4-ID7), or by increasing R at constant 𝜆𝑝 (Figure 4a-b against 

Figure 4f, i.e. ID1-ID2 against ID9). On the other hand, increasing R had a negligible effect on the fuel bed 

temperatures for high air excess ratios, as also shown by Figure 3, because of a large amount of cold air injected 

into the system (see Figure 4e, i.e. ID6). Figure 4 shows that different thermal regions appear in the solid 

biomass zone from the left to the right area of the fixed bed. 

On the bottom-left side of the fuel bed, lower temperatures were found possibly due to the high moisture content 

of the biomass; a drying phase near the inlet region appeared due to the fresh fuel supply. In the central area, 

the higher temperature indicated that combustion reactions, both in gas and solid phase, took place. In all the 

cases, the bottom-right side of the fuel bed temperatures were lower than 600 °C. This was due to the 

accumulation of depleted fuel promoted by the motion of the screw feeder, which pushed fresh biomass from 

the bottom-left to the bottom-right of the system. This fact was already observed in other recent studies involving 

the same device in different operating conditions (Caposciutti et al., 2018). Figure 4 also highlights that the EGR 

can drastically affect the extent of these zones. In particular, large gas recirculation pushed the drying zone 

towards the inlet position, resulting in a more uniform temperature profile in the remaining part of the fixed bed.  

4. Conclusions 

Exhaust gas recirculation was investigated in a 140 kW biomass underfed stoker boiler to analyse its impact on 

the temperatures in the combustion region. In particular, the temperature distribution inside and above the fixed 

bed was analyzed as a function of the air excess and recirculation ratios. When the recirculation was introduced, 

the behaviour of the average bed temperature versus the air excess ratio changed. Without recirculation, in 

facts, an optimal value of the air excess exists, as a result of the interplay between better mixing and fresh air 

dilution. When the recirculation was increased, this trend was smoothed and the average temperature trend with 

the excess ratio resulted more uniform. The recirculation ratio also affects the behaviour of the fixed bed local 

temperatures. For similar air excess ratios, a higher rate of recirculated flue gas increases the average biomass 

bed temperatures and leads to a more even temperature distribution. Moreover, it pushes the main reaction 

area towards the fuel inlet zone, drastically affecting the thermal zones distribution in the bed of combusting 

biomass. These details can be useful for developing solid combustion lumped-parameter models, and for 

validating CFD models. For instance, the present data may support the development of in-bed models based 

on reactor networks to be coupled to freeboard CFD calculations. 

References 

Archan G., Anca-Couce A., Buchmayr M., Hochenauer C., Gruber J., Scharler R., 2021, Experimental 

Evaluation of Primary Measures for NOX and Dust Emission Reduction in a Novel 200 kW Multi-Fuel 

Biomass Boiler, Renewable Energy, 170, 1186–96. 

Caposciutti G., Barontini F., Antonelli M., Tognotti L., Desideri U., 2018, Experimental Investigation on the Air 

Excess and Air Displacement Influence on Early Stage and Complete Combustion Gaseous Emissions of a 

Small Scale Fixed Bed Biomass Boiler, Applied Energy, 216, 576–87. 

Caposciutti G., Barontini F., Galletti C., Antonelli M. Tognotti L., Desideri U., 2020, Woodchip Size Effect on 

Combustion Temperatures and Volatiles in a Small-Scale Fixed Bed Biomass Boiler, Renewable Energy, 

151, 161-174. 

Carroll J.P., Finnan J.M, Biedermann F., Brunner T., Obernberger I., 2015, Air Staging to Reduce Emissions 

from Energy Crop Combustion in Small Scale Applications, Fuel 155, 37–43. 

Gómez M.A., Martín R., Chapela S., Porteiro J., 2019, Steady CFD Combustion Modeling for Biomass Boilers: 

An Application to the Study of the Exhaust Gas Recirculation Performance, Energy Conversion and 

Management, 179, 91–103. 

Lamberg H., Sippula O., Tissari J., Jokiniemi J., 2011, Effects of Air Staging and Load on Fine-Particle and 

Gaseous Emissions from a Small-Scale Pellet Boiler, Energy & Fuels, 25, 4952–60. 

Patronelli S., Antonelli M., Tognotti L., Galletti C., 2018, Combustion of Wood-Chips in a Small-Scale Fixed-Bed 

Boiler: Validation of the Numerical Model through in-Flame Measurements, Fuel, 221, 128–37. 

Rajh B., Yin C., Samec N., Hriberšek M., Kokalj F., Zadravec, M., 2018, Advanced CFD Modelling of Air and 

Recycled Flue Gas Staging in a Waste Wood-Fired Grate Boiler for Higher Combustion Efficiency and 

Greater Environmental Benefits, Journal of Environmental Management, 218, 200–208. 

Schulte S., Heinzle J., Groß B., 2020, Reduction of Particulate Matter Emissions from a Pellet Boiler Using 

Primary Measures, Chemical Engineering & Technology, 43, 1506–13. 

 

402


	150caposciutti.pdf
	An Experimental Investigation on the Effect of Exhaust Gas Recirculation in a Small-Scale Fixed Bed Biomass Boiler