Atmosphere

Emissions from domestic solid fuel burning appliances (wood-heaters, open fireplaces)

Technical Report No. 5
J. Gras, C.Meyer, I. Weeks, R. Gillett, I. Galbally, J. Todd, F. Carnovale, R. Joynt, A. Hinwood, H. Berko and S. Brown.
Environment Australia, March 2002
ISBN 0 6425 4867 6

6. Time variation of particle mass emission rates

Up to this point, emissions have been described in terms of emission factors, that is the amount of emission per quantity of fuel consumed, integrated over a complete burn. This gives no appreciation for the rate at which material is emitted nor for which stage of the burn the material is emitted. Several continuous measurements were made that give this information for the aerosol emissions and also some of the gas components. Measurements related to aerosol mass included a TEOM continuous mass balance, a TSI DustTrak and a Radiance M903 nephelometer. All of these instruments were fed from the secondary dilution system on the aerosol sample line (see Fig. 2). The DustTrak is a compact instrument that determines mass loading indirectly by light scattering. In this respect it is similar to a nephelometer and, as was expected, these two instruments generally tracked each other very closely, albeit with different nominal output scales. The nephelometer is calibrated in terms of volume scattering coefficient whereas the DustTrak reports using a nominal mass-loading scale based on test dusts referred to as Arizona Road Dust. An example of the output from all three of these instruments is shown in Fig. 32. In this figure output from the DustTrak and TEOM are plotted on the same scale and scattering coefficient, determined with the nephelometer, has an independent scale. Whilst the TEOM is a useful instrument for ambient measurements of aerosol mass loading, it proved to be unsuitable for this study. The response time is too slow (see Fig. 32), and the tendency to lose volatile and semi-volatile material is a problem. An example of this can be seen in Fig. 32, where negative mass loadings are evident following the end of the emission peak, at around 12:15. There is also a substantial difference in indicated mass loadings from the DustTrak and TEOM. In part this can be attributed to the loss of semi-volatile mass from the TEOM. TEOM data will not be discussed further since these are not considered to be a good representation of the actual emissions.

Figure 32: Time sequence of output from the TEOM, DustTrak (nominal mass loading) and nephelometer (scattering coefficient) for sample #17, run on 15th Feb.

Figure 32: Time sequence of output from the TEOM, DustTrak (nominal mass loading) and nephelometer (scattering coefficient) for sample #17, run on 15th Feb.

The data that will be used to describe the particle mass emissions were determined using the DustTrak, with normalisation to the observed gravimetric mass for each sample. The DustTrak and nephelometer gave output that was effectively equivalent; hence the nephelometer data will not be discussed further.

6.1. Time series of mass emission rates

The time course of mass emission rates is given in Figs. 33–40, and represents the output from the DustTrak normalised or calibrated so that the integral of the time series is identical to the observed gravimetric mass collected for each determination. Units used for the emission rate are g per kg of fuel charge, per hour. In this and subsequent discussions the mass of fuel considered is the dry fuel load. Data are grouped by test type, with the first group, dry eucalypts burnt in the controlled-combustion heaters with the flow set to maximum, given in Fig. 33. Most of the mass emissions occur over a relatively short period, typically 20 minutes or less, commencing immediately after the loading of the fuel. With bluegum the emission peak was slightly delayed although it still occurred within the 20-minute window. All but one of these samples were taken using the AS4013-compliant heater C2.

Figure 33: Variation of aerosol mass emission with time, in minutes after loading the heater, for eucalypt and high flow setting.

Figure 33: Variation of aerosol mass emission with time, in minutes after loading the heater, for eucalypt and high flow setting.

Note: C2 = AS4013-compliant heater (nominal 3.7 g/kg), NC = non-compliant heater, RG = redgum, JA = jarrah, BG = bluegum #1 = sample 1 etc.

The time course for emissions from dry redgum at the lowest flow setting, shown in Fig. 34, was determined for all three of the controlled-combustion heaters and showed some differences between these appliances. Heater C2 typically returned a prolonged emission lasting around 2 hours whereas the non-compliant heater and the low-emission heater C1 gave significant mass emissions for about 20 minutes, similar to the previous high flow-rate burns on C2. Differences in the time-course for these three heaters at the flow settings used, suggests that only C2 had a significant change in air supply over the maximum range of settings. For low flow tests a 10% fuel burn-off with the maximum flow setting was used to establish combustion before adjusting the air supply to the low setting. This burn-off typically took 5 ±1 min. The initial peak, which can be seen in Fig. 34, results from quenching when the air-flow control is moved from the high to low settings.

Figure 34: Variation of aerosol mass emission with time, in minutes after loading the heater, for dry redgum (water < 16%) and low flow setting.

Figure 34: Variation of aerosol mass emission with time, in minutes after loading the heater, for dry redgum (water < 16%) and low flow setting.

Note: C1 = AS4013-compliant heater (nominal 0.9 g/kg), C2 = AS4013-compliant heater (nominal 3.7 g/kg), NC = non-compliant heater, #3 = sample 3 etc.

The open fireplace insert, an uncontrolled combustion device, behaved as expected, similarly to a controlled-combustion heater with a high flow setting. This is shown in Fig. 35.

The time series for particle mass emissions from very wet redgum (water content typically 34% of fuel weight) was determined for heater C2 at both low and high flow settings. At the high flow setting, the time course of the emissions from this fuel was comparable to that for seasoned eucalypt with a short, less than 20-minute peak of similar magnitude to those for seasoned fuel (Fig. 36). At low flow settings however, the pattern was quite different. After the initial 20% fuel burn-off which was used to establish combustion, and which took approximately 21 minutes, selection of the low flow setting effectively quenched the emissions which then very slowly redeveloped over a two to three hour period to a value typically around 5 g/kg/h.

Figure 35: Variation of aerosol mass emission with time, in minutes after loading heater, for open fireplace insert burning dry redgum.

Figure 35: Variation of aerosol mass emission with time, in minutes after loading heater, for open fireplace insert burning dry redgum.

Note: water =14.9%, #6 = sample 6 etc.

Figure 36: Variation of aerosol mass emission with time, in minutes after loading the heater, for AS4013-compliant heater C2 burning green redgum.

Figure 36: Variation of aerosol mass emission with time, in minutes after loading the heater, for AS4013-compliant heater C2 burning green redgum.

Note: water = 33.8%

Overnight burns with a grossly overloaded heater and the lowest flow setting from the start of the burn (no initial burn-off) were conducted on heater C2 and also the non-compliant heater. Mass emission rates from these burns are given in Fig. 37. For heater C2 this resulted in an emission time-course similar to that for the very fresh green eucalypt (Fig. 36), but remaining more-or-less uniformly around 5 g/kg/h for around 2½ hours and falling away after 3 hours. The non-compliant heater also produced a relatively uniform emission time course, but in this case lasting only around 40 to 50 minutes and more typically around 8 or 9 g/kg/h. This is consistent with a significantly higher flow rate through the non-compliant heater on the low flow setting. Given the age and condition of the heater, relatively poor performance is not surprising.

Figure 37: Variation of aerosol mass emission with time, in minutes after loading the heater, for AS4013-compliant heater C2 and the non-compliant heater burning dry redgum with a grossly overloaded combustion chamber.

Figure 37: Variation of aerosol mass emission with time, in minutes after loading the heater, for AS4013-compliant heater C2 and the non-compliant heater burning dry redgum with a grossly overloaded combustion chamber.

Note: water = 15.1%–15.5%

Manufactured fuel was tested only in heater C2 with a high flow setting. The time course of the emissions for this fuel, shown in Fig. 38, differed slightly from that for eucalypts with similar combustion conditions, being delayed about 10 minutes and with a longer period of emission, at around 30 minutes. This is plausibly a result of the relatively high density of this fuel, 1070 kg/m³.

Figure 38: Variation of aerosol mass emission with time, in minutes after loading the heater, for AS4013-compliant heater C2 burning manufactured fuel (extruded logs) on a high flow rate.

Figure 38: Variation of aerosol mass emission with time, in minutes after loading the heater, for AS4013-compliant heater C2 burning manufactured fuel (extruded logs) on a high flow rate.

The final group of time-series is given in Figs. 39 and 40, representing emissions from the combustion of pine in heater C2. For dry pine on the high flow setting, the mass emissions were intense, lasting 15 to 20 minutes and with a magnitude some ten times those from eucalypt. For the low setting, the initial burn-off on the high setting was to 10% fuel reduction. Complete quenching did not occur on reduction of the flow although the emissions were initially significantly reduced and then decayed steadily over the subsequent 20-minute period (see Fig. 39). For the very green unseasoned pine (water content 54% of fuel mass), a 20% burn-off was employed to enable sustained combustion. This occurred over the first 10 to 15 minutes. Quenching of the flame occurred on reduction of the flow setting, resulting in a long smoldering period persisting for more than four hours (Fig. 40). The dried pine that was re-wet by immersion in water for 45 minutes also showed delay in establishing combustion with the 20% initial fuel burn-off requiring around 10 minutes. Following reduction of the air supply to minimum setting, there was initial quenching and a subsequent decay of emissions over about a one-hour period (see Fig. 39). This pattern is intermediate between emissions from the dried and very green fuels. Evidently, heaters such as C2, that are designed for burning hardwood, are inappropriate for burning dry pine, and can result in intense bursts of particle emissions whenever new dry fuel is added.

Figure 39: Variation of aerosol mass emission with time, after loading heater, for AS4013-compliant heater C2 burning pine at the indicated moisture content and flow rates.

Figure 39: Variation of aerosol mass emission with time, after loading heater, for AS4013-compliant heater C2 burning pine at the indicated moisture content and flow rates.

Figure 40: Same as Fig. 39 but with extended time scale and expanded emission factor scale.

Figure 40: Same as Fig. 39 but with extended time scale and expanded emission factor scale.

6.2 Time variation of mass emission rates of gaseous pollutants

Combustion of wood in a controlled-combustion heater is a complex, even chaotic process. There are many competing and complementary processes, dependent on the physical and chemical state of the fuel, heater design, temperature, state of the combustion etc, see for example Shelton (1983), Ballard-Tremeer (1997), Skreiberg (1997). Sequences of emissions in this section illustrate some of this complexity. Time series for emissions of the gas species, VOC, CO, CO2, NOX and SO2 for representative cases that include burns with low and high fuel moisture contents and high and low airflow conditions, are shown. This includes an example of an 'overnight' burn with an overloaded combustion chamber and low flow rate setting.

For redgum burns emission rates are plotted in Figs. 41–45, as a function of time after loading the fuel charge, and for pine in Figs. 49–53. Additional data on the state of the combustion for these burns, comprising power output, combustion efficiency (C[CO2]/ΣC), flue temperature and CO to CO2 ratio are plotted in Figs. 46–48 for the redgum burns and in Figs. 54–55 for the pine burns. Emission rates for VOC, CO and CO2 are plotted as g per species per kg of fuel per hour whilst those for NOX and SO2 are plotted as mg per species per kg of fuel per hour.

For the high flow rate burns using both redgum and pine, all of the species considered here (VOC, CO, CO2, NOX and SO2) show a very rapid rise in emission rate immediately following the loading of a new fuel charge onto the coal bed (see Figs. 41–45, 49–53). Maximum emission rates are reached relatively early in the burn in all cases although there are clear differences between species with SO2 production peaking first, followed by VOC and CO, then CO2 and NOX later. For all species, the maximum emission rate was reached much earlier for the pine burns. For the high flow setting burns there is an initial drop in flue temperature when the fuel is first inserted, but as is evident in Figs. 46, 47 and 48, the temperature begins to rise immediately. A relatively short period of smoldering is evident from the CO/CO2 ratio until widespread pyrolysis occurs. Pyrolysis is the thermal breakdown of cellulose and hemicellulose into volatiles and char with flaming combustion the burning of the volatiles, typically with a luminous yellowish flame. Pyrolysis is the source of the evolved VOCs, with unburnt VOCs and high CO concentrations indicative of oxygen limitation. Most aerosol mass emissions occur concurrently with the VOCs as condensates of the heavier volatile fraction.

NOX emission rates tend not to closely follow the initial pyrolysis-induced unburnt fuel peak shown, for example by VOC or particle mass, but rather the chamber temperature or power output (see Figs. 46, 47, 54), with maximum emission rates broadly coincident with maximum flue temperatures. Detailed studies of nitrogen combustion chemistry in controlled fluidised bed combustion chambers (Skreiberg et al., 1997; Winter et al., 1999) confirm that production of NO occurs predominantly during the later stages of devolatilisation, when combustion temperatures are high (700 °C–900 °C, optimally near 850 °C) and oxygen partial pressure is near 15 kPa. At lower temperatures and oxygen concentration, the formation of reduced nitrogen compounds (nitriles, HCN and NH3) is favoured. Nitrous oxide production is always very much less than nitric oxide production (at most 4% of fuel nitrogen) and occurs predominantly between flame extinction and char combustion in the presence of higher oxygen partial pressure (15–21 kPa). In our study high combustion temperature and high oxygen concentration did not occur simultaneously.

For all of the emissions, the time to the peak in emission rate was shortest for dry pine, followed by dry redgum and then green redgum. The relativity reflects the rate that heat penetrates fuel of differing density and the delaying effect of fuel moisture which insulates the fuel interior and requires energy input to evaporate (densities for the different fuel types are indicated in Table 16). Elimination of fuel moisture during establishment of pyrolysis also contributes to reduced oxygen supply through dilution.

Table 16: Density of fuel types used, expressed on a dry weight basis.
  density kg/m³ (dry basis)
seasoned redgum 890
green redgum 730
bluegum 780
jarrah 640
pine 410
manufactured logs 1070

With combustion of the volatiles the chamber temperature and power output rises, thermally driven convection increases the oxygen supply and the concentration of CO falls. Overall combustion efficiency increases to a broad maximum (see for example Figs. 46, 47 and 54). For these burns, in the latter stage of combustion after elimination of the volatiles, a glowing combustion phase where the fire largely burns char was established. A soft bluish flame marks this form of combustion. Only towards the end of the burn as the temperature and convective draught fall, oxygen depletion is evident and CO/CO2 again rises.

Figure 41: Emission of VOCs from redgum combustion in heater C2.

Figure 41: Emission of VOCs from redgum combustion in heater C2.

Note: Variables include sample number (#), fuel moisture (%), airflow and loading (standard or overloaded).

Figure 42: Emission of NO from redgum combustion in heater C2.

Figure 42: Emission of NO from redgum combustion in heater C2.

Note: Variables include sample number (#), fuel moisture (%), airflow and loading (standard or overloaded).

Figure 43: Emission of CO2 from redgum combustion in heater C2.

Figure 43: Emission of CO2 from redgum combustion in heater C2.

Note: Variables include sample number (#), fuel moisture (%), airflow and loading (standard or overloaded).

Figure 44: Emission of NOX from redgum combustion in heater C2.

Figure 44: Emission of NOX from redgum combustion in heater C2.

Note: Variables include sample number (#), fuel moisture (%), airflow and loading (standard or overloaded).

Figure 45: Emission of SO2 from redgum combustion in heater C2.

Figure 45: Emission of SO2 from redgum combustion in heater C2.

Note: Variables include sample number (#), fuel moisture (%), airflow and loading (standard or overloaded).

Figure 46: Combustion efficiency, flue temperature, power output and CO/CO2 for dry redgum, heater C2, high and low flow settings.

Sample #37: 12.8% water, high flow

Figure 46a: Combustion efficiency, flue temperature, power output and CO/CO2 for dry redgum, heater C2 - sample 37

Sample #35: 12.8% water, low flow

Figure 46b: Combustion efficiency, flue temperature, power output and CO/CO2 for dry redgum, heater C2 - sample 35

Figure 47: Combustion efficiency, flue temperature, power output and CO/CO2 for green redgum, heater C2, high and low flow settings.

Sample #39: 33.8% water, high flow

Figure 47a: Combustion efficiency, flue temperature, power output and CO/CO2 for green redgum, heater C2 - sample 39

Sample #32: 33.8% water, low flow

Figure 47b: Combustion efficiency, flue temperature, power output and CO/CO2 for green redgum, heater C2 - sample 32

Figure 48: ombustion efficiency, flue temperature, power output and CO/CO2 for green redgum, heater C2, with overloaded combustion chamber, sample #16

Figure 48: ombustion efficiency, flue temperature, power output and CO/CO2 for green redgum, heater C2, with overloaded combustion chamber, sample #16

Figure 49: Emission of VOCs from pine combustion in heater C2.

Figure 49: Emission of VOCs from pine combustion in heater C2.

Note: Variables include sample number (#), fuel moisture (%), airflow and loading (standard or overloaded).

Figure 50: Emission of CO from pine combustion in heater C2.

Figure 50: Emission of CO from pine combustion in heater C2.

Note: Variables include sample number (#), fuel moisture (%), airflow and loading (standard or overloaded).

Figure 51: Emission of CO2 from pine combustion in heater C2.

Figure 51: Emission of CO2 from pine combustion in heater C2.

Note: Variables include sample number (#), fuel moisture (%), airflow and loading (standard or overloaded).

Figure 52: Emission of NOX from pine combustion in heater C2.

Figure 52: Emission of NOX from pine combustion in heater C2.

Note: Variables include sample number (#), fuel moisture (%), airflow and loading (standard or overloaded).

Figure 53: Emission of SO2 from pine combustion in heater C2.

Figure 53: Emission of SO2 from pine combustion in heater C2.

Note: Variables include sample number (#), fuel moisture (%), airflow and loading (standard or overloaded).

Figure 54: Combustion efficiency, flue temperature, power output and CO/CO2 for pine, heater C2, high and low flow settings.

Sample 45: 11.3% water, high flow

Figure 54a: Combustion efficiency, flue temperature, power output and CO/CO2 for pine, heater C2 - sample 45

Sample 51: 12.3% water, low flow

Figure 54b: Combustion efficiency, flue temperature, power output and CO/CO2 for pine, heater C2 - sample 51

Figure 55: ombustion efficiency, flue temperature, power output and CO/CO2 for pine, heater C2, sample #50, 54.3% water, low flow.

Figure 55: ombustion efficiency, flue temperature, power output and CO/CO2 for pine, heater C2, sample #50, 54.3% water, low flow.

For the low flow-rate burns, the time sequences observed for the combustion and emissions were significantly different to those for the high flow rate burns. Two examples for AS4013-like loading of the heater with pine and two with redgum, illustrating burns with dry and green fuel, are provided. This group also includes an example simulating an overnight burn, with a grossly overloaded combustion chamber, dry redgum fuel and a low airflow setting.

In all cases the fuel was placed into a hot combustion chamber. For the dry redgum burn on low flow (#35) the initial burn-off was 10% and for the overloaded combustion chamber burn (#16) there was no initial burn-off on high flow setting. This resulted in a very early onset of strong smoldering, identifiable in Figs. 46 and 48 as the high CO/CO2 ratio and low combustion efficiency. During this phase there was a small amount of flame for the standard-loading low burn (sample #35) and little, if any, visible flame for the overloaded burn (#16). For both, the temperature and power output fell steadily. After about one hour the CO/CO2 ratio started to decline and combustion efficiency improved, reaching a peak after about 1.5 hours for the standard-loading burn and 4 hours for the overloaded case. For the overloaded chamber burn, particle mass emissions remained high during this period with low efficiency smoldering, for around three hours (Fig. 37), as did VOC and SO2 concentrations (Figs. 41 and 44). For the AS4013-like load of dry redgum burnt on low setting, the period of elevated emissions was shorter, around one hour, but also corresponding to the period of smoldering combustion (see Figs. 34, 41).

For very green redgum on a low flow setting (sample #32), 20% of the fuel was burnt off on high setting to establish sustained combustion, after which the flow was set to low. This induced strong smoldering as shown by the high CO/CO2 ratio, which peaked after one hour (see Fig. 47). Only small flames were present in the initial burn-off period and after closing to the low flow setting there were no visible flames. Combustion subsequently improved, relatively steadily, for the next two hours. During this extended smoldering period, particle mass emissions, VOCs and SO2 remained at an extended high level (see Figs. 36, 41–45).

As with the high flow cases, the NOX emission rates tend to reflect the chamber temperature or power output (see Figs. 44, 47), with maximum emission rates broadly coincident with maximum flue temperatures. The magnitude of the emission rate is also much lower for these lower temperature burns, whether due to low flow, fuel loading or fuel moisture content.

Comparing the pine burns, with the hardwood burns (Figs. 49–53 and Figs. 41–48), both fuels have similar time-course patterns but the combustion of pine was far more rapid producing very much stronger emissions of most components and particularly NOx, VOCs, CH4, CO and CO2. This was true for both the high flow and low flow burns and for dry and green fuel.

The variation in NOX emission rates for the low setting pine burns also reflects the chamber temperature or power output (see Figs. 54 and 55). In these cases maximum emissions occurred in the hotter parts of the burn immediately after the fuel was added during the initial fuel burn-off on high flow while the chamber temperature was still elevated, with a second small emission peak later when combustion conditions improved.

6.3 Particle Size distribution and changes in the distribution with time

In both epidemiological studies and legislative control related to health effects of airborne particles, most emphasis has been placed on measures related to the atmospheric mass loading. For some physiological processes, such as macrophage response, the number of particles, and particularly very small particles, may also play an important role (see, for example, Gras, 1996 pp.116–117). Particle number concentrations and sizes in smoke, like mass concentrations, vary with different combustion regimes although it is known that virtually all smoke particles usually lie within the respirable range (Daero < 10 mm) and a clear majority ( > 90%) have a diameter smaller than 1 µm.

For this study, in general, the particle number concentrations were greatest during the initial phase of high particulate mass emissions that occur after loading the fuel charge and before the flue temperature recovers to around 300 °C, conditions indicative of the commencement of widespread pyrolysis. This is shown in Figs. 56–58 for some selected examples representing conditions for overloaded dry redgum (15.5% water), green redgum (33.8% water) and dry pine (12% water). In these figures, cumulative concentrations in the dilution tunnel, for particles larger than three selected diameters, are given. These show both the gross variation of particle number concentration and give some indication of size distribution changes as a function of time.

Figure 56: Aerosol particle concentration for AS4013-compliant heater C2, dry redgum (water = 15.5%), grossly overloaded combustion chamber and low flow (sample #16).

Figure 56: Aerosol particle concentration for AS4013-compliant heater C2, dry redgum (water = 15.5%), grossly overloaded combustion chamber and low flow (sample #16).

Note: Indicated diameters (D) in mm.

Figure 57: Aerosol particle concentration for AS4013-compliant heater C2, green redgum (water = 33.8%), high flow (sample #39).

Figure 57: Aerosol particle concentration for AS4013-compliant heater C2, green redgum (water = 33.8%), high flow (sample #39).

Note: Indicated diameters (D) in mm.

Figure 58: Aerosol particle concentration for AS4013-compliant heater C2, dry pine (water = 12%), high flow (sample #49).

Figure 58: Aerosol particle concentration for AS4013-compliant heater C2, dry pine (water = 12%), high flow (sample #49).

Note: Indicated diameters (D) in mm.

A better measure of the changes in size distribution can be gauged from Figs. 59–61, where differential size distributions giving the particle concentration by logarithmic size interval are plotted. These distributions, which represent conditions in the dilution tunnel, have the log-normal pattern typically observed for particles emitted from combustion sources. In all cases, the distributions show a systematic change with time after the fuel load is added to the heater. Initially, particle size is relatively large with the geometric mean diameter near or exceeding 0.2 µm. This corresponds to the initial period before pyrolysis is fully established, with low combustion chamber temperature, poor combustion efficiency and high mass emissions. These particles are expected to contain a high proportion of unburnt condensates distilled from the fuel during the pre-pyrolysis phase. As the fire progresses the size of particles emitted decreases, with the geometric mean diameter in the later stages of the burn more typically around 0.05 µm. The rate of the decrease in size is dependent on the type of burn. Establishment of the smaller mode size coincides with the fall in mass emission rate at the onset of pyrolysis. Note that times for distributions refer to the end of a determination, and also that for some plots the pre-loading distribution is shown for reference (e.g minute -4 in Fig. 59 and 0 in Fig. 60).

Figure 59: Aerosol size distributions for AS4013-compliant heater C2, dry redgum (water = 15.5%), grossly overloaded combustion chamber and low flow (sample #16).

Figure 59: Aerosol size distributions for AS4013-compliant heater C2, dry redgum (water = 15.5%), grossly overloaded combustion chamber and low flow (sample #16).

Note: Values in the legend are minutes after loading fuel.

Figure 60: Aerosol size distributions for AS4013-compliant heater C2, green redgum (water = 33.8%), high flow (sample #39).

Figure 60: Aerosol size distributions for AS4013-compliant heater C2, green redgum (water = 33.8%), high flow (sample #39).

Note: Values in the legend are minutes after loading fuel.

Figure 61: Aerosol size distributions for AS4013-compliant heater C2, dry pine (water = 12%), high flow (sample #49).

Figure 61: Aerosol size distributions for AS4013-compliant heater C2, dry pine (water = 12%), high flow (sample #49).

Note: Values in the legend are minutes after loading fuel.