DOI: 10.3303/CET2292084 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
Paper Received: 24 January 2022; Revised: 13 March 2022; Accepted: 13 May 2022 
Please cite this article as: Bergomi A., Morreale C., Fermo P., Migliavacca G., 2022, Determination of Pollutant Emissions from Wood-fired 
Pizza Ovens, Chemical Engineering Transactions, 92, 499-504  DOI:10.3303/CET2292084 
  

 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 

Determination of Pollutant Emissions from Wood-Fired Pizza 

Ovens 

Andrea Bergomia, Carmen Morrealeb, Paola Fermoa, Gabriele Migliavaccab,* 

aUniversità degli Studi di Milano, Dipartimento di Chimica, Via Golgi 19, 20133, Milano  
bInnovhub Stazioni Sperimentali per l’Industria, Area Combustibili, Via Galilei 1, 20097, San Donato Milanese (MI) 

gabriele.migliavacca@mi.camcom.it 

The present study deals with the determination of pollutant emissions from wood-fired pizza ovens throughout 

the implementation of an innovative sampling system, specifically designed for the source in question. The 

experiments were conducted on a traditional wood-fired oven using two different forms of beechwood as fuel: 

wood logs and briquettes. Three different combustion cycles were used to carry out the tests, two for the wood 

logs and one for the briquettes. Different sampling sites along the system allowed for the collection and analysis 

of several gaseous pollutants, such as nitrogen oxides (NOx), carbon monoxide (CO), carbon dioxide (CO2) and 

OGC (organic gaseous carbon); along with total particulate matter (TPM). In order to study the synergistic effect 

of biomass burning and pizza cooking, combustion cycles were performed both with and without the presence 

of the pizzas. The results show that the concentrations of gaseous pollutants were affected mainly by the 

combustion conditions in the oven, whereas the values of TPM depended more on the amount of fuelwood 

used. Pizza cooking had a great impact on the emissions of OGC and, to a lesser extent, on CO and TPM.  

1. Introduction 

Non-industrial combustion is one of the eleven categories of emitters identified by the European Environmental 

Agency (EEA) listed in the Selected Nomenclature for Air Pollution (SNAP) (Dietmar, 1998). This macrosector 

includes combustion processes aimed at the production of heat for non-industrial activities such as commercial 

and institutional plants, residential heating and domestic combustion processes such as fireplaces and stoves, 

and agricultural installations (ARPA Sicilia, 2018). Non-industrial combustion contributes to the emissions of all 

major air pollutants (SO2, NOx, CH4, CO, CO2, N2O, NH3 and volatile organic compounds), with a significant 

impact on particulate matter. For instance, according to the most recent emission inventories developed by the 

INEMAR database, 42.5% of all the PM10 emitted in the Lombardy region of Italy can be traced back to non-

industrial combustion, and this percentage reaches 49.1% if PM2.5 is considered (ARPA Lombardia, 2018).  

Correlations between emissions from the civil sector and the different fuels used highlight the fact that wood 

combustion is the most relevant source of all major pollutants, with the exception of nitrogen oxides, in which 

case fossil fuels contribute the most (ARPA Sicilia, 2018). From the analysis of specific tracers such as 

levoglucosan and benzo(a)pyrene, several studies identified the rural mountainous areas as the most influenced 

by wood combustion; however, important contributions from biomass burning have been found also in urban 

areas (Lanzani, 2014). Moreover, the problem of air pollution deriving from wood burning is not confined to 

specific regions, but is widespread around the world, impacting mainly on the levels of PM2.5 (Favez et.al., 

2009) and PM10 (Borrego et.al., 2010). 

Wood burning in the non-industrial sector is employed mainly in small-scale domestic heating and cooking 

appliances including closed and open fireplaces, traditional and advanced stoves, boilers and kitchen stoves 

(Ozgen et.al., 2014), along with wood-burning ovens used in pizzerias and bakeries (Ielpo et.al., 2020). Whereas 

the contribution in rural mountainous areas is dominated by small-scale appliances used for heating purposes 

(Lanzani, 2014), in urban areas there is a non-negligible contribution deriving from wood-burning ovens (ARPA 

Sicilia, 2018). To the knowledge of the authors, there have been very few studies concerned with calculating 

the extent of this contribution. Within the Metropizza project (Hugony, 2018), a first attempt was made starting 

from emission inventory data using a modelling approach; the results showed that wood-burning pizza ovens in 

499



the province of Milan contributed in particular to the emissions of particulate matter and polycyclic aromatic 

hydrocarbons (PAHs). 

Despite these preliminary studies, there is still an impressive lack of experimental data on the emissions of this 

source. To date, only one study has been published dealing with the characterization of particles emitted by 

ovens burning wood and briquettes. The results from this work highlight that wood-burning pizza ovens are an 

important source of fine particulate matter (PM2.5) characterized by a high content of black carbon (20-30%) 

(Mota Lima et.al., 2020). Moreover, there are no specific technical standards defining the constructional aspects 

of sampling systems or specifying tests methods for the analysis of pollutants. The only standard document that 

can deal with the subject is a 1995 UNI standard, currently withdrawn (UNI 10474:95), which deals with solid 

fuel ovens. In the broader field of biomass combustion, the European Commission has introduced source-

specific emission standards in the Ecodesign Directive (EDD) for other appliances such as woodstoves and 

fireplaces, however excluding wood-fired ovens (Wolters, 2018).  

The present work is a preliminary study on the emission profile of wood-fired ovens used in pizzerias. The aim 

is to determine the concentrations of the main air pollutants emitted by this source and to evaluate the extent to 

which pizza cooking impacts the overall emissions. In order to address these issues, an innovative sampling 

system was designed enabling the collection and analysis of the desired pollutants, and two different types of 

fuels were tested. Moreover, combustion cycles were created with the aim of reproducing the real-life operating 

conditions of the oven and at the same time ensuring repeatability and reproducibility of the measurements.  

2. Materials and Methods 

2.1 Appliance and fuels 

The experimental tests were carried out using a traditional fixed top wood-fired pizza oven, which included a 

cooking surface made of refractory slabs with high-percentage alumina. A large amount of refractory material 

was also present under these plates, guaranteeing a supply of heat without temperature changes. Also, the 

cooking chambers were made of vibrated refractory concrete.  

The oven was fed with beech wood logs normally sold to pizzerias, of length 50 cm, split diameter 6-8 cm, and 

with a moisture content below 20%. Instead, the wood briquettes were made of compressed beech wood without 

any use of adhesives or additives, certified for food use with a moisture content between 5.4-6.6%. The length 

of the briquettes was 30 cm and the diameter was 8.5 cm (octagonal shape). 

2.2 Sampling system 

The sampling system employed in the execution of experimental tests on the oven was constructed by making 

reference to the UNI EN 16510-1:2019 technical standard, which describes the test methods for sampling and 

measurement of the pollutant emissions of solid biomass residential heating appliances (Figure 1).  

As reported in Annex F of the aforementioned standard, this system enables the simultaneous measurement of 

pollutants and other substances on the hot flue gases and on the cold flue gases after dilution in the tunnel. The 

dilution is controlled by a variable power exhaust fan located at the top of the dilution tunnel, and through a 

damper on a secondary section of the dilution tunnel.  

The gaseous substances (CO, NOx, CO2, O2) were measured both on the concentrated and diluted fumes, 

except for OGC which was measured only before dilution. Instead, TPM was sampled in continuous mode on 

the diluted gases using an opacimeter, and in discontinuous mode using appropriate filters on both the hot and 

cold flue gases. Also, the draught in the flue gas duct was continuously measured on the hot gases, and adjusted 

by varying the distance between the top of the measuring section and the suction hood of the dilution tunnel, as 

well as by varying the settings of the exhaust fan by means of an inverter.  

2.3 Dilution factor 

In order to report the concentrations of TPM measured in the diluted flue gases under the same conditions, it is 

necessary to determine the degree of dilution achieved in the tunnel. This was done by using the measured 

concentration values of CO2 both upstream [CO2upstream] and downstream [CO2downstream] of dilution and 

subsequently applying the following formula to calculate the dilution factor, d, which takes into account also the 

concentration of carbon dioxide in ambient air, which was 300 ppm.      

 (1) 

Due to the fact that the dilution factor depends on the flue gas flow rate instantaneously generated by the 

combustion, as well as on the initial tunnel settings, this parameter may vary from test to test and during each 

test. It was therefore calculated punctually on the bases of the concentrations averaged over every 

measurement (every 10 seconds).   

𝑑 = ([𝐶𝑂2𝑢𝑝𝑠𝑡𝑟𝑒𝑎𝑚 ] − 0.03)/([𝐶𝑂2𝑑𝑜𝑤𝑛 𝑠𝑡𝑟𝑒𝑎𝑚 ] − 0.03) 

500



 

 

Figure 1: Scheme of the sampling system (not to scale)  

2.4 Combustion cycles 

In order to study the emissions of residential heating appliances burning solid biomass technical standards have 

laid out the combustion cycles that need to be followed (EN 13240:2001, EN 13229:2001, EN 14785:2006). Due 

to the lack of regulations regarding wood-fired ovens, new combustion cycles were developed specifically for 

the purposes of this study. Tests on wood-fired ovens must be conducted in such a way as to simulate, as far 

as possible, the real-life operating conditions of an oven in a normal public establishment. This is highly 

dependent on specific and local factors, such as the type of establishment and clientele, the season and local 

habits. However, it is necessary to define a standard reference cycle to be used in conducting the tests.  

First of all, the oven used in the tests was managed following five operational phases: 

 

1. Cold ignition: the cold furnace is ignited after several days of dormancy with a load of wood used as kindling, 

sold on the market specifically for this purpose; 

2. Pre-heating: a series of fuel loads are selected for the tests and are fed to the furnace in sequence, taking 

advantage of the bed of embers formed during the ignition phase, in order to bring the furnace to operating 

temperature (approximately 350 °C); 

3. Hot re-ignition: the furnace, which has been inactive for several hours (12-14), is re-ignited, as in point 1, 

when its internal temperature is still relatively high (80-180 °C), as is normally the case following the daily 

opening of a business, and is then fed with successive fuel loads until it reaches the operating temperature. At 

that point, the fuelwood and the embers are moved to one side of the oven, leaving space to cook the pizzas 

on the other side in the following phase; 

4. Cooking phase: the hot oven is periodically fed with fuel loads in order to maintain a constant operating 

temperature, and at the same time, the pizzas are introduced. The normal operations of cleaning and moving 

the ashes and embers that are necessary for effective management of the oven under operating conditions are 

also carried out, but are excluded from the combustion cycles.     

5. Shutdown: during this phase no fuel is supplied and no firing operations are carried out.  

 

The pollutant emissions of the oven were measured during the cooking phase. The combustion cycles defined 

consisted of a single load of fuelwood, sampled for a fixed amount of time starting from the moment of entry of 

the fuel in the oven. Depending on the type of fuel and on the amount of beechwood used for the tests, different 

modifications were made to the standard reference cycle (Table 1).  

 

501



Table 1: Combustion cycle details 

Combustion cycle Fuelwood Mass of fuelwood supplied for each 

load / kg 

Sampling time / 

minutes 

A Beechwood logs 1.7-1.9 20  

B Beechwood logs 0.8-1.0 15 

C  Beechwood briquettes 1.7 15 

 

In order to evaluate the impact of pizza cooking on the overall emissions, during each day of sampling 4 cycles 

were conducted with pizza cooking and 3-4 cycles were carried out without pizza cooking, for a total of 7-8 loads 

for each test on a particular fuel. The start of the first cycle of every day was set once the temperature of the 

oven reached 350°C; the others followed the first one.  

2.5 Sampling methods and techniques 

All the instrumentation used during the execution of the tests complies with the indications of the UNI EN 16510-

1 technical standard and is managed in accordance with the specific reference technical standards for each 

parameter whose concentration is determined in the combustion fumes: UNI EN 14789:2017 for O2, UNI EN 

15058:2017 for CO, UNI EN 14792:2017 for NOx, UNI EN 12619:2013 for OGCs, and UNI EN 13284-1:2017 

for TPM.  

All of the gaseous pollutants were measured using continuous analysers which were calibrated before the start 

of each test using calibration mixtures of known concentration. The fumes were extracted using a heated probe 

at 180°C and conditioned using a ceramic filter heated at 200°C, along with a Chilly compressor refrigeration 

unit and internal pump with dew-point output of 5°C.  

Instead, smoke sampling for the determination of TPM was carried out on a section of the emission duct, as 

established by the reference standard, using a probe with a housing at the end in which the sampling filter was 

positioned; downstream of the probe are placed in succession: a smoke cooling device consisting of a 

condensate collector immersed in an ice bath, an absorption tower with silica gel and the suction pump.  

The mass concentration of the emitted dust was determined gravimetrically after proper conditioning. The filters 

subjected to hot sampling were initially placed in an oven at 180°C for 1 hour, then in a desiccator for 4 hours, 

again in the oven at 160°C for one hour, and finally in the desiccator for another 4 hours before being weighed. 

Instead, the ones subjected to cold sampling were left to dry for 24 hours in a desiccator before being weighed. 

Finally, the opacimeter used to sample TPM continuously on the diluted fumes was not calibrated: the absolute 

values do not reflect the actual dust concentration; the only purpose of this measurement was to highlight 

possible trends during the combustion cycles.  

3. Results and Discussion 

3.1 Pollutant emissions of the wood-fired oven 

The main parameters kept under control for a correct execution of the test were: the draught, which significantly 

influences the combustion process and therefore the operation of the oven; the dilution factor, and the 

temperature of the diluted smoke at the sampling point, which affects the degree of condensation of semi-volatile 

organic compounds (SVOCs). Average values are reported in Table 2, along with other key sampling 

parameters.  

Table 2: Sampling parameters (average values) 

Combustion 

Cycle  

Wood consumption / 

kg h-1  

Hot flue gas 

temperature / °C 

Cold flue gas 

temperature / °C 

Draught / Pa Dilution factor 

A 6.2 189 54.4 -18 5.4 

B 5.1 153 44.1 -19 5.2 

C 5.8 174 51.6 -16 4.8 

 

Another important sampling parameter is the number of pizzas cooked during each day of tests: this was kept 

constant at a value equal to 16.  

The emissions of the air pollutants for the various cycles were calculated as the average of the measured values 

for each cycle, with and without pizza cooking (Table 3). The impact of the pizzas will be discussed in the 

following section. NOx and TPM concentrations fall within the wide range of values observed in literature data 

for domestic wood-burning appliances (Tomlin, 2021). Differently, when compared to the same devices, the 

502



tested oven emitted lower concentrations of CO and OGC, closer to the performances of pellet fuelled 

appliances (Ozgen, 2014).  

Table 3: Average concentrations of the measured substances 

 HOT FLUE GASES  COLD FLUE GASES 

Cycle NOx / 

ppm 

CO / 

ppm  

CO2 /         

%v 

O2 / 

%v 

OGC /   

mg m-3 
TPM /           

mg Nm-3 
 NOx / 

ppm 

CO / 

ppm  

CO2 / 

%v 

O2 /  

%v 

Dust 

(opacimeter) / 

mg m-3 

TPM /           

mg Nm-3 

A 21.8 165 2.8 18.4 10.5 100 (±10)  4.3 34.2 0.54 20.5 30.0 91 (±8) 

B 15.0 355 1.9 19.4 26.3 24 (±4)  2.8 71.8 0.39 20.8 6.0 30 (±3) 

C 18.5 262 2.9 18.3 16.3 96 (±11)  4.2 59.3 0.66 20.3 36.1 105 (±9) 

 

These results show that the concentration of CO and OGC emitted by the oven are greater when performing 

loadings using smaller amounts of beechwood separated by shorter time intervals (Cycle B). An opposite trend 

is observed in terms of NOx and CO2: these pollutants are emitted in greater amounts when performing loadings 

with greater amounts of beechwood logs separated by longer time intervals (Cycle A). This is consistent with 

the fact that emissions from biomass combustion are highly dependent on operational practices (Van Loo and 

Koppejan, 2008). Regarding Cycle C, the gaseous emissions of the oven were generally half-way between 

Cycle A and Cycle B, with the exception of CO2, in which case the values were closer to Cycle A. 

Based on experimental observations, loading more beechwood (Cycles A, C) enabled to reach higher average 

temperatures in the oven and keep a high flame throughout the entire sampling period. Instead, tests using 

Cycle B were associated with a reduction of the flame’s intensity during the final sampling minutes, which caused 

the levels of CO and OGC to rise before the end of each cycle as a result of poor combustion conditions. Instead, 

no significant differences were observed in the trends of NOx and CO2 in all the aforementioned cycles. However, 

unlike other studies (Li et.al., 2012), higher temperatures were linked to greater emissions of NOx. 

With regards to TPM, Cycles A and C produced greater amounts of dust compared to Cycle B. This suggests 

that the amount of particulate matter generated is strictly related to the amount of fuelwood used. Overall, the 

comparison between TPM sampled on the hot and cold flue gases does not highlight a significant degree of 

condensation of semi-volatile organic compounds (SVOCs). On the one hand, this is in line with the findings of 

Ozgen et.al. (2014) which highlight a much lower difference between TPM concentrations on the hot and cold 

flue gases for open systems (open fireplaces, ovens) as opposed to closed ones (stoves, boilers). On the other 

hand, given that OGC was detected at concentrations above 10 mg m-3 in all three cycles, further measurements 

will be carried out to clarify this observation.  

3.2 The impact of pizza cooking 

Table 4 shows the average results obtained for experimental tests conducted with and without pizza cooking for 

the different cycles.  

Table 4: Average concentrations with and without pizza cooking 

   HOT FLUE GASES  COLD FLUE GASES 

Cycle Pizza  NOx / 

ppm 

CO / 

ppm  

CO2 /         

%v 

O2 / 

%v 

OGC /   

mg m-3 
TPM /           

mg Nm-3 
 NOx / 

ppm 

CO / 

ppm  

CO2 

/         

%v 

O2 /  

%v 

Dust 

(opacimeter) / 

mg m-3 

TPM /           

mg Nm-3 

A 

yes  22.4 182 2.9 18.3 13.7 109 (±16)  4.4 36.2  0.6 20.5 32.5 98 (±11) 

no  20.9 145 2.7 18.4 6.3 87 (±8)  4.2 31.4 0.5 20.5 26.6 75 (±6) 

B 
yes  15.5 354 2.0 19.2 35.3 31 (±1)  2.8 71.4 0.4 20.7 6.5 28 (±2) 

no  14.6 357 1.8 19.4 17.4 17 (±7)  2.9 72.3 0.4 20.8 5.5 26 (±5) 

C 
yes  18.6 284 2.9 18.3 24.3 93 (±16)  3.9 61.7 0.6 20.3 32.8 110 (±15) 

no  18.4 225 2.9 18.3 7.5 99 (±15)  4.1 50.2 0.6 20.3 29.5 100 (±11) 

 

In terms of nitrogen oxides and carbon dioxide, no significant differences can be observed between the tests 

with and without pizza cooking. Differently, Cycles A and C suggest that this activity had an impact on the 

emissions of CO, whereas the results obtained for Cycle B do not show this trend, possibly due to large 

503



increases observed at the end of each trial which concealed the effect induced by the presence of the pizzas. 

Instead, the results from all three cycles agree on the fact that pizza cooking had a significant impact on the 

emissions of OGC. However, the gaseous compounds emitted from the pizzas did not have an impact on 

particulate matter. In fact, as already seen for the average results in the previous section, no significant amount 

of condensation can be observed in any one of the tests and the concentrations of TPM emitted with and without 

pizza cooking are not statistically different. One possible explanation could be that the expected contribution of 

the condensable fraction related to the higher OGC emissions associated with pizza cooking is concealed by 

the very high variability of the TPM measurements associated with wood burning. Once again, this problem will 

be further addressed in the following stages of this study.  

4. Conclusions 

From the results obtained in this study it is possible to conclude that wood-fired ovens represent an important 

source of major air pollutants. The concentrations of the examined species were comparable with the 

performances of other wood- or pellet-burning small-scale appliances. The experimental tests also showed that 

pizza cooking had a significant impact on the emission of OGC as a result of the volatilization of several 

compounds present in the food matrix. 

Due to their extensive presence in many urban areas, the need to introduce regulation for this sector is of the 

upmost importance for the safeguard and improvement of air quality. At the European level, several actions 

have already been put in place in order to reduce the impacts of biomass combustion in the domestic sector, 

which could possibly be extended to include wood-fired ovens.  

The next stage of the project will include the sampling and analysis of other air pollutants, such as polycyclic 

aromatic hydrocarbons (PAHs), with the aim to achieve a complete emission profile of this source. In addition, 

the contribution of PM2.5 and PM10 to TPM will be determined, along with a complete speciation of particulate 

matter into its various constituents: OC, EC, heavy metals, anions and cations. Furthermore, the emissions 

during ignition and shutdown of the oven will be established and compared with the ones determined during the 

cooking phase.    

References 

ARPA Lombardia, 2018, INEMAR - INventario EMissioni in AtmosfeRa: Emissioni in Regione Lombardia 

nell'anno 2014 - Dati finali, ARPA Lombardia Settore Monitoraggi Ambientali, IT.  

ARPA Sicilia, 2018, Piano Regionale di Tutela della Qualità dell’Aria in Sicilia, Report, Regione Siciliana, IT. 

Dietmar K., 1998, Annual Topic Update, Report, European Topic Centre on Air Emissions, Copenhagen, DK. 

Borrego C., Valente J., Carvalho A., Sà E., Lopes M., Miranda A.I., 2010, Contribution of residential wood 

combustion to PM10 levels in Portugal, Atmospheric Environment, 44, 642-651. 

Favez O., Cachier H., Sciare J., Sarda R., Martinon L., 2009, Evidence for a significant contribution of wood 

burning aerosols to PM2.5 during the winter season in Paris, Atmospheric Environment, 43, 3640-3644. 

Hugony F., 2020, Progetto Metropizza, Seminar, Città metropolitana di Milano – ENEA, Milano, IT.  

Ielpo, P., Placentino, C.M., Genga, A., Ancona, V., Uricchio, V.F., Fermo, P., 2020, PM2.5 in indoor air of a 

bakery: Chemical characterization and size distribution, Atmosphere, 11, 4, 415.  

Lanzani G., 2014, La combustione domestica delle biomasse legnose e qualità dell’aria, Conference 

Presentation, Conferenza del sistema nazionale per la protezione dell'ambiente, Bologna, IT.  

Li Z., Liu Y., Chen Z., Zhu Q., Jia J., Li J., Wang Z., Qin Y., 2012. Effect of the air temperature on combustion 

characteristics and NOx emissions from a 0.5 MW pulverized coal-fired furnace with deep air staging, Energy 

Fuels, 26, 4, 2068–2074. 

Mota Lima F.D., Pérez-Martìnez P.J., de Fatima Andrade M., Kumar P., de Miranda R.M., 2020, 

Characterization of particles emitted by pizzerias burning wood and briquettes: a case study at Sao Paulo, 

Brazil, Environmental Science and Pollution Research, 27, 35875-35888. 

Ozgen S., Caserini S., Galante S., Giugliano M., Angelino E., Marongiu A., Hugony F., Migliavacca G., Morreale 

C., 2014, Emission factors from small scale appliances burning wood and pellets, Atmospheric Environment, 

94, 144-153. 

Tomlin A. S., 2021, Air Quality and Climate Impacts of Biomass Use as an Energy Source: A Review, Energy & 

Fuels 35, 18, 14213–14240. 

Van Loo S., Koppejan J., 2008. The Handbook of Biomass Combustion and Co-firing. Earthscan, London. 

Wolters R., 2018, EU policy regarding emission reduction from domestic combustion, Report, European 

Commission. 

504


	186bergomi.pdf
	Determination of Pollutant Emissions from Wood-Fired Pizza Ovens