Operating Conditions of a Portuguese Wood Stove

Operating conditions of a Portuguese wood stove

A.I. Calvo, L.A.C. Tarelho, M. Duarte, M. Evtyugina, C.A. Alves

Centre for Environmental and Marine Studies (CESAM), Department of Environment and Planning, University of Aveiro, 3810-193 Aveiro, Portugal

+351 234 370 200, [email protected]

1. Introduction – Wood is commonly used in residential combustion for heating. As it is a renewable energy source, its use is recommended given that enables to reduce dependency on fossil fuels. However, wood combustion is considered as an important source of air pollutants, with a significant environmental and health impact. In the last years, an especial attention has been point out in biomass burning emissions, focussing on emissions from wildfires, prescribed fires and agricultural burning. However, an important fraction of all biomass combustion occurs in domestic stoves and fireplaces that, although of small scale, are used in considerable number, having important potential to contribute to the emission of particles and gases [1]. Puxbaum et al. [2] argued that emissions from residential wood combustion could be responsible for up to 70% of the organic particulate matter in the atmosphere of European rural locations during winter. The stove design, operating conditions, combustion conditions (e.g. amount of excess air) and the species of wood and their characteristics are factors that can have an important influence on pollutant emissions. Incomplete combustion is among the main problems in domestic equipments that largely contribute to the emission levels observed; incomplete combustion not only influence local air quality but also exerts some influence at global level because of the emission of greenhouse gases [1]. Improved stoves offer the potential to increase the efficiency of fuel conversion, decrease the negative impact on public health and reduce adverse environmental impact [3]. In addition, economic issues can’t be neglected, as operating conditions are related to energetic and environmental performance that meet international standards in order to be commercialised worldwide. In Portugal, fireplaces and wood stoves are very common for space heating at homes. However, the information about the operating conditions of batch operated stoves with typical Portuguese biomass fuels is scarce. The main goal of this study is to characterise the operating conditions and the flue gas emissions from combustion of wood from two common Portuguese trees ( Eucalyptus globulus (hardwood) and Pinus pinaster (softwood)) in a stove. Improving knowledge on operating conditions of these heating equipments and on composition of flue gas is an important challenge considering the achievement of efficient, economic and environmental compatible combustion.

2. Experimental – The biomass combustion experiments were carried out with a typical Portuguese stove, commonly used for domestic heating (Fig. 1). The stove is characterised by a combustion chamber with a volume of 0.093 m 3, corresponding to 0.44 m height, 0.59 m width and 0.36 m depth. It was equipped with a vertical chimney with 0.2 m internal diameter and 3.3 m height. Wood from two typical Portuguese tree species, softwood (Pinus pinaster ) and hardwood (Eucalyptus globulus ), was used as fuel (Table 1). The wood was cut into logs of 0.3 to 0.4 m in length with a total biomass burned during each cycle of around 1.7 to 2.0 kg. Combustion cycles lasted from 45 to 60 minutes. The combustion air enters the combustion chamber below the stove grate, thus primary air, and flows throughout the grate and the fixed bed of burning biomass. The driving force for the air entering the combustion chamber is natural convection resulting from the up flowing column of hot combustion flue gases throughout the vertical chimney. The combustion flue gas composition was monitored continuously at the exit of the chimney, namely for Total Volatile Hydrocarbons (THC) (Dyna-FID Hidrocarbon Gas Analyser, model SE-310), O2 (paramagnetic, ADC model O 2-700 with a Servomex Module), CO 2 and CO (non-dispersive infrared, Environnement MIR 9000). Temperature was monitored continuously using two K-type thermocouples at different locations: i) in the stove combustion chamber, around the centre of the cross section of the combustion chamber at about 0.20 m above the grate (T3 in Fig. 1), and ii) in the chimney at 2.8 m above the exit of the stove combustion chamber (T6 in Fig. 1). Continuous monitoring of the weight of the fuel at the grate of the stove and of the air flow rate entering the combustion chamber was performed using a Combustion flue gases Heated sampling line (190ºC) Cold water (in) weight sensor (load cell DSEUROPE Warm water (out)

H Sample Gas (SG) / Pressure Model 535QD-A5) and a mass flow G Termocouple T6 meter (Kurz, Model: 500-2.0-P 40),

G respectively.

(

)

/ For O , CO and CO monitoring the 2 2

r e gas sampling and characterisation I UCD1 Zero Gas (ZG) system included a water-cooled

N2 045

000 F HC N SG SG UCD0 sampling probe, a set of gas J O O2 L ZG conditioning and distribution units, and SG

h h P t t CO2 Termocouple T5 M a a

B B a set of on-line gas analysers (Fig. 1).

e e

c c

I I K CO Q UCD2 The combustion flue gas was sampled Condensed -1 material at a flow rate of 2 L min (atmospheric Pressure R UCE1

E Temperature pressure and temperature). S For total volatile hydrocarbons (THC) Termocouple T4 analysis, expressed as CH 4, a heated probe and sampling line (at 190ºC)

A Termocouple T3 was used for gas sampling; the heated B Termocouple T2 sampling line delivers the flue gas Termocouple T1 sample directly to the FID analyser.

C D Air The combustion flue gas was sampled for O 2, CO, CO 2 and THC in the chimney at 2.8 m above the exit of the Figure 1 . Schematic representation of the experimental installation. Dashed stove combustion chamber (Fig. 1). line – Electric circuit, Continuous line – Pneumatic circuit. A - Stove, B – Grate of the stove, C – Load cell (weight sensor), D – Air flow meter, E – Thermal insulation of the exhaust duct, F – Exhaust duct (Chimney), G – 3. Results and Discussion – A set of Water-cooled gas sampling probe, H – Heated sampling line, I, J, K – experiments (3 for pine and 4 for Command and gas distribution units (UCD0, UCD1, UCD2), L – Gas eucalyptus) was carried out in order to sampling pump, M – Gas condensation unit for moisture removal, N, O, P, Q – Automatic on-line gas analysers (THC, CO 2, O 2, CO), R – Electronic determine the operating conditions and command unit (UCE1), S – Computer data acquisition and control system. the flue gas characteristics from combustion of pine and eucalyptus wood. The results obtained during the continuous monitoring of the stove operation include information about: i) fuel weight in the grate, ii) air flow rate entering the combustion chamber, iii) temperature of the flue gas in the combustion chamber and at the chimney exit, and iv) O 2, CO 2, CO and THC concentration in the flue gas at the exit of the chimney (Fig. 2). The lack of data at 30 minutes of experiment is related to the need to stop and restart the data acquisition system. The general trend observed during the combustion of the two wood types was similar, and is characterised by three main stages: i) a first stage corresponding to the initial heating accompanied by wood drying and the initial steps of devolatilisation without the existence of a visible flame, ii) a second stage characterised by the devolatilisation, ignition and combustion of volatiles and char during which it is observed a vigorous flame, and iii) a third stage characterised by the combustion of the char, during which there is almost no visible flame. The analysis of the data below will be focused on the behaviour of the two specific combustion experiments presented in Fig. 2, but for the other experiments the general trends are similar. The combustion air was atmospheric air and the flow rate ranged between 27 and 39 Nm 3/h. For both fuels, the combustion air flow rate increased suddenly in the first 5-10 minutes, reaching a maximum value, and then decreasing gradually until the end of the combustion experiment (Fig. 2). The temperature in the combustion chamber increases suddenly in the first 5 minutes and reaches values as high as 650ºC at around 15 minutes after loading a batch of fuel. There is a direct relationship between the temperature in the combustion chamber and the temperature of the flue gas at the exit of the chimney (Fig. 2).

Table 1 . Characteristics of the biomass fuel Moisture Ash C H N S O (by difference) Biomass %wt, fuel %wt, fuel dry basis as received Pinus pinaster 9.7 0.46 51.40 6.20 0.16 nd 41.88 Eucalyptus globulus 10.0 0.25 48.60 6.20 0.16 nd 44.28 nd – not determined because the concentration was below the equipment detection level of 0.1%wt. a1) a2) 2000 50.00 2000 50 /h) . /h) 3 3 air 40.00 40 1500 1500 air

30.00 30 1000 1000 20.00 20 mass mass 500 500 10.00 10 Mass of fuel in the grate (g) grate the in Massfuel of Mass of fuel in the grate (g) grate the in Massfuel of Combustion air flow rate (Nm rate flow air Combustion 0 0.00 (Nm rate flow air Combustion 0 0 0:00:00 0:15:00 0:30:00 0:45:00 1:00:00 0:00:00 0:15:00 0:30:00 0:45:00 1:00:00 Time (hh:mm:ss) Time (hh:mm:ss) b1) b2) 1000 1000

800 800

T T3 600 3 600

400 400 T6 T6 Temperature(ºC) Temperature (ºC) 200 200

0 0 0:00:00 0:15:00 0:30:00 0:45:00 1:00:00 0:00:00 0:15:00 0:30:00 0:45:00 1:00:00 Time (hh:mm:ss) Time (hh:mm:ss) c1) c1) 800 800

600 600

400 400

200 200 Temperature in exit flue gas (ºC) flue exit in Temperature Temperature in exit flue gas (ºC) flue exit in Temperature 0 0 0 200 400 600 800 0 200 400 600 800 Temperature in combustion chamber (ºC) Temperature in combustion chamber (ºC) d1) d2) 21 8000 21 8000

18 O2 18 O2 6000 6000 15 15

12 12 4000 4000 9 CO 2 9 CO (%v, dry gases) (%v, dry gases)

2 CO 2 CO 2

, O , 6 , O ,

2 6 2000 2 2000 CO(ppmv, dry gases) CO(ppmv, dry gases) CO 3 CO CO 2 3 CO 2 CO CO 0 0 0 0 0:00:00 0:15:00 0:30:00 0:45:00 1:00:00 0:00:00 0:15:00 0:30:00 0:45:00 1:00:00 Time (hh:mm:ss) Time (hh:mm:ss) e1) e2) 5000 5000

4000 4000 , ppmv, wet wet ppmv, , 4 , ppmv, wet ppmv, , 4 3000 3000

gases) 2000 gases) 2000

1000 1000

THC(expressed as CH 0 THC(expressed as CH 0:00:00 0:15:00 0:30:00 0:45:00 1:00:00 0 0:00:00 0:15:00 0:30:00 0:45:00 1:00:00 Time (hh:mm:ss) Time (hh:mm:ss)

Figure 2 . Evolution of operating variables during the combustion of a batch of fuel for pine (1) and eucalyptus (2): a) fuel weight in the grate, air flow rate entering the combustion chamber, b) temperature of the flue gas in the combustion chamber (at T3, in Fig. 1) and at the chimney exit (at T6, in Fig. 1), c) relationship between the temperature of the flue gas at the chimney exit (at T6, in Fig. 1) and the temperature in the combustion chamber (at T3, in Fig. 1) of the wood stove d) O 2, CO 2, CO (G in Fig. 1), and e) total volatile hydrocarbons concentration in the flue gas at the chimney exit (H in Fig. 1). 3.5 The temperature at the top of the chimney reached Pinus pinaster

) 3 Eucalyptus globulus values up to 260ºC and 280ºC, for pine and -1 eucalyptus, respectively. 2.5 The biomass consumption along the experiment shows 2 two distinct phases (Fig. 2): i) a first phase, during the 1.5 first 30 min, with average rate of solid fuel conversion -1 1 of around 3 kg h , and ii) a second phase, from 30 min

Rate of fuel conversion (kg h (kg conversion fuel of Rate till the end of the experiment, with average rate of 0.5 solid fuel conversion lower than 1 kg h-1 (Fig. 3). The 0 average rate of fuel conversion during the entire Start - 30 min 30 min - End Start - End -1 Figure 3 . Rate of fuel conversion (kg h-1) during the first combustion cycle for both fuels is around 2 kg h 30 min (Start-30 min), from 30 minutes to the end of the (Fig. 3), with pine presenting slight higher average experiment (30 min-End), and for the complete burning consumption rates. experiment (Start-End), for pine and eucalyptus. A direct relationship between temperature in the combustion chamber and combustion air flow rate was observed; a high combustion air flow rate admitted to the combustion chamber is induced by high temperatures in the combustion chamber, as a consequence of a higher natural convection driving force induced by the hot combustion gases throughout the chimney. During the initial 15 minutes of the combustion of a biomass batch a concentration peak of CO and sudden decrease on the concentration of O 2 are observed, followed by an increase on the concentration of CO 2 in the combustion flue gases. This process is accompanied by a sharp increase on the temperature in the combustion chamber, and corresponds to a time period characterised by relatively high solid fuel consumption rate (Figure 2). Afterwards, the concentration of O 2 starts to increase slowly, whereas that of CO 2 decreases, till the end of the combustion of the fuel batch. After around 30 minutes from the beginning of the experiment, that is, during the combustion of the char, the CO concentration starts to increase gradually, reaching relatively high concentrations (up to 4000 ppmv, dry gases) despite the existence of a high concentration of O 2 (> 17%v); this could be associated with the relatively low temperature (<500ºC) and residence time in the combustion chamber. The significant influence of temperature on the oxidation of CO has been previously recognised [4]. In the final stage of combustion, beyond the 40 minutes, the CO concentration starts to decrease gradually till the end of the experiment. It is worth noting the significant concentrations of CO registered during the last phase of the combustion experiment, mainly due to healthy reason given that human poisoning by CO inhalation as a result of exposure to gases emitted from wood burning in old fireplaces and stoves in poorly ventilated places has long been recognised. The THC concentration profile along the combustion cycle follows generally the same trend as the CO concentration. At the initial stage of combustion it is observed a concentration peak, which can reach values as high as 3700 ppmv (wet gases), and coincident with the initial phase of wood devolatilisation. At around 30 minutes of experiment, that is after the vigorous flame extinction, and during the combustion of char the THC shows a slight increase followed by a smooth decrease till the end of the combustion experiment.

4. Conclusions – The operating conditions of a Portuguese wood stove were characterised, considering fuel consumption rate, flue gas temperature and composition. The initial stage of combustion of a batch of fuel is critical, because the rapid devolatilisation of the solid originate relatively high amounts of CO and hydrocarbon species in a short time scale, and this compounds are not oxidised despite the high concentration of O 2 available; this could be related to relatively low temperatures and insufficient residence time of the gases in the combustion chamber, associated to deficient gas mixing. A significant CO concentration in the flue gas was observed during the final phase of combustion in which no flame is visible and mainly burning of char takes place. Improvements on such equipments are needed in order to achieve a more environment friendly conversion of biomass to energy.

5. References [1] J. Zhang, K.R. Smith, Y. Ma, S. Ye, F. Jiang, W. Qi, P. Liu, M.A.K. Khalil, R.A. Rsmussen, S.A. Thorneloe, 2000. Atmos. Environ . 34 , 4537-4549. [2] H. Puxbaum, A. Caseiro, A. Sánchez-Ochoa. M. Clayes, A. Gelencsér, M. Legrand, S. Preunkert, C. Pio, 2007. J. Geophys. Res . 11 2, D23S05. [3] Ezzati M., Mbinda B.M., and Kammen D.M., 2000. Environ. Sci. Technol. 34 (4), 578–583. [4] E.M. Lipsky, A.L. Robinson, 2006. Environ. Sci. Technol. 40 , 155-162.