Effect of Moisture on Gas Emissions from Stored Woody Biomass

Biomass materials have been increasingly used due to their renewable nature. The problems occurring during the storage of fresh woody materials include gas emissions and dry matter losses as a result of degradation. The objective of this study was to investigate and quantify the effect of moisture content on gas emissions from stored wood chips. Experiments were conducted under non-aerobic and aerobic conditions using fresh Western Red Cedar (WRC) chips with different initial moisture contents over a range of temperatures. The peak CO2 emission factor of 2.9 g/kg dry matter (DM) was observed from high moisture chips at 20 °C under non-aerobic conditions after two-month storage, which was an order of magnitude greater than that from low moisture chips. In the case of volatile organic compounds, a range of compounds were detected from all tests. The concentration of VOCs was found to be positively correlated with moisture content. Gas emissions from the aerobic reactors exhibited similar trends as non-aerobic reactors with respect to the effect of moisture content, although higher values were observed under aerobic conditions. Slight reduction of dry mass from all tests at the end of storage indicated the decay-resistance characteristics of WRC.


Introduction
There is a growing demand for bioenergy in today's world. Biomass as a renewable resource form is becoming an important element of sustainability and more attractive to the public due to its abundance and reduced air pollution in terms of criteria air pollutants and greenhouse gases as compared to most types of fossil fuels [1]. In particular, a large amount of woody biomass (forest residues) can be harvested annually in North America to produce heat and power as well as a rather broad range of products that include biofuels, pulp and paper, panels, chemicals and other energy feedstocks [2].
The immediate use of woody biomass after harvest is generally infeasible. Depending on the production cycles, the biomass is stored in the field for some time before transportation and processing. It is particularly necessary to store the biomass materials for an extended time period during the winter months in order to ensure their uninterrupted supply and delivery to the conversion facilities. The need for woody fuels storage, especially wood chips, is also in line with the increasing demand for bioenergy.
Nevertheless, the prolonged storage of waste biomass prior to utilization may cause some problems. The physicochemical and biological degradations of the biomass during storage will lead to gas emissions that have impact on the surrounding environment and human health [3][4][5]. Gas emissions from stored biomass will also result in a loss of dry matter and the heating value of the loss. In particular, emissions from stored WRC chips with different moisture contents have not been studied. The objective of this paper is to investigate the gas emission from stored WRC chips with different moisture contents and its effect on dry matter loss under different storage conditions. Findings from this study will be useful in developing safe storage and transportation of woody biomass, and contribute towards assessing the potential impact of gas emission on the surrounding environment and human health.

Materials
Western red cedar chips used in this study were obtained from Langley, BC, Canada ( Figure 1). The cedar logs originating from Vancouver Island were transported in water booms to sawmills in the Lower Mainland Vancouver area. The chips with a uniform size for pulp or bioenergy applications were produced from waste wood which was cut off to make commercial products. The chips were stored in a room at 4 • C at UBC laboratory before the experiment. The average size of the chips as measured by the Gilson Testing Screens (Gilson Company, Worthington, OH, USA) had a range of 10-20 mm in length, 5-10 mm in width and 1-3 mm in thickness. The moisture content of the wood chips was determined in triplicate in a forced-air convection oven at 103 • C for 24 hours according to ASABE Standards S358.2 [33]. The initial moisture content of the chips was about 50% (wet basis). The wood chips were further dried in an oven to obtain the moisture content around 20% (wet basis) and 35% (wet basis), respectively. The moisture content of 35% simulates the value of naturally dried wood chips, while the value of 20% is after artificial drying. All materials were then stored at 4 • C before the test.
be one source of biomass feedstock for densification via pelletizing as some two-thirds of the wood pellets manufactured in North America (Canada and USA) are shipped to Europe and Asia [32].
Based on the literature review, moisture content is an important factor that affects the gas emissions from stored woody biomass. However, the impact of moisture content on stored wood chips has not received much attention, especially the extent of its influence on gas emission and dry matter loss. In particular, emissions from stored WRC chips with different moisture contents have not been studied. The objective of this paper is to investigate the gas emission from stored WRC chips with different moisture contents and its effect on dry matter loss under different storage conditions. Findings from this study will be useful in developing safe storage and transportation of woody biomass, and contribute towards assessing the potential impact of gas emission on the surrounding environment and human health.

Materials
Western red cedar chips used in this study were obtained from Langley, BC, Canada ( Figure 1). The cedar logs originating from Vancouver Island were transported in water booms to sawmills in the Lower Mainland Vancouver area. The chips with a uniform size for pulp or bioenergy applications were produced from waste wood which was cut off to make commercial products. The chips were stored in a room at 4 °C at UBC laboratory before the experiment. The average size of the chips as measured by the Gilson Testing Screens (Gilson Company, Worthington, OH, USA) had a range of 10-20 mm in length, 5-10 mm in width and 1-3 mm in thickness. The moisture content of the wood chips was determined in triplicate in a forced-air convection oven at 103 °C for 24 hours according to ASABE Standards S358.2 [33]. The initial moisture content of the chips was about 50% (wet basis). The wood chips were further dried in an oven to obtain the moisture content around 20% (wet basis) and 35% (wet basis), respectively. The moisture content of 35% simulates the value of naturally dried wood chips, while the value of 20% is after artificial drying. All materials were then stored at 4 °C before the test.

Experimental Setup
Wood chips with different initial moisture contents (20%, 35% and 50%, wet basis) were divided into three series in this study. Series 1 to 3 represented the WRC chips with initial moisture content of 20%, 35% and 50%, respectively. For each series, materials were separated to two groups, one group under aerobic condition and the other group under non-aerobic condition. This is to simulate the different environmental conditions that the materials in a pile experience during storage depending on the availability of oxygen.
A number of 2 L glass containers were used to conduct the storage test. The containers were fitted with valves and sampling ports. Each assembled reactor was then loaded with 330 g wood chips. As shown in Figure 2, non-aerobic reactors were sealed at all times to study gas emissions

Experimental Setup
Wood chips with different initial moisture contents (20%, 35% and 50%, wet basis) were divided into three series in this study. Series 1 to 3 represented the WRC chips with initial moisture content of 20%, 35% and 50%, respectively. For each series, materials were separated to two groups, one group under aerobic condition and the other group under non-aerobic condition. This is to simulate the different environmental conditions that the materials in a pile experience during storage depending on the availability of oxygen.
A number of 2 L glass containers were used to conduct the storage test. The containers were fitted with valves and sampling ports. Each assembled reactor was then loaded with 330 g wood chips. As shown in Figure 2, non-aerobic reactors were sealed at all times to study gas emissions under airtight conditions. After the daily gas sampling, air was pumped into the aerobic reactors in order to replenish and maintain a high oxygen level in the reactors. under airtight conditions. After the daily gas sampling, air was pumped into the aerobic reactors in order to replenish and maintain a high oxygen level in the reactors. After all the reactors were loaded with the materials, they were sealed and placed under different temperatures at 5 °C, 20 °C, 35 °C and 50 °C. This range of storage temperature represents cool to hot climate conditions in different geographic locations, and involving seasonal variations. Two replicates were performed for each test. The average values of the results obtained from the two replicates were then presented. In all cases, the experiment period was approximately two months.

Gas Emission Measurements
The emissions of CO2 and CO from the tests were analyzed by gas chromatography (Model SRI8610 C, Mandel, Torrance, CA, USA). The packed column GC was equipped with FID/TCD detector, and using argon as carrier gas. The oven was heated up to 400 °C with a retention time of 15 min. The GC was calibrated regularly with the corresponding standard gases. During each sampling event, gas sample was drawn by using syringe from each reactor to measure the concentrations of CO2, CO and O2 concentrations. Gas emissions from the aerobic reactors were measured daily. Whereas, gas emissions from the non-aerobic reactors were measured more frequently at the beginning of the experiment, and every few days in the later period of the test. 250 μL gas sample was drawn from each reactor at the end of the experiment to analyze the NMVOC compositions by GC/MS analyzer (Model 5975B/6890N, Agilent Technologies, Santa Clara, CA, USA). The GC was equipped with capillary column and FID detector, and using helium as carrier gas. The oven was heated to 50 °C, and then ramped 20 °C/min till 200 °C for 3 min.
The total concentrations of NMVOCs (TVOC) from the aerobic reactors were measured by a portable VOC monitor (Model PGM-7240, RAE Systems, San Jose, CA, USA) on a daily basis. The ppbRAE photoionization detector was employed to measure the concentration by plugging the detection end to the reactor outlet after daily opening. The measured values were recorded after stabilization. For the reactors under non-aerobic conditions, the same procedure cannot be applied since this would cause a larger amount of gases to be released from the reactor, thus affecting the accuracy of gas analysis for the remaining test (storage) period. Thus, TVOC concentrations associated with the non-aerobic reactors were measured at the end of the test period.

Data Analysis
In this study, the emissions of gases are expressed as emission factors, in units of [gram gas species per kilogram dry matter DM]. Emission factor is a cumulative parameter. The measured gas concentrations in vol% were converted to emission factors by using the N2 balance method [13], assuming that nitrogen was not consumed during the test period. At constant temperature and pressure, the emission factor f in [g/kg DM] is related to volumetric gas concentration Ci as follows: After all the reactors were loaded with the materials, they were sealed and placed under different temperatures at 5 • C, 20 • C, 35 • C and 50 • C. This range of storage temperature represents cool to hot climate conditions in different geographic locations, and involving seasonal variations. Two replicates were performed for each test. The average values of the results obtained from the two replicates were then presented. In all cases, the experiment period was approximately two months.

Gas Emission Measurements
The emissions of CO 2 and CO from the tests were analyzed by gas chromatography (Model SRI8610 C, Mandel, Torrance, CA, USA). The packed column GC was equipped with FID/TCD detector, and using argon as carrier gas. The oven was heated up to 400 • C with a retention time of 15 min. The GC was calibrated regularly with the corresponding standard gases. During each sampling event, gas sample was drawn by using syringe from each reactor to measure the concentrations of CO 2 , CO and O 2 concentrations. Gas emissions from the aerobic reactors were measured daily. Whereas, gas emissions from the non-aerobic reactors were measured more frequently at the beginning of the experiment, and every few days in the later period of the test. 250 µL gas sample was drawn from each reactor at the end of the experiment to analyze the NMVOC compositions by GC/MS analyzer (Model 5975B/6890N, Agilent Technologies, Santa Clara, CA, USA). The GC was equipped with capillary column and FID detector, and using helium as carrier gas. The oven was heated to 50 • C, and then ramped 20 • C/min till 200 • C for 3 min.
The total concentrations of NMVOCs (TVOC) from the aerobic reactors were measured by a portable VOC monitor (Model PGM-7240, RAE Systems, San Jose, CA, USA) on a daily basis. The ppbRAE photoionization detector was employed to measure the concentration by plugging the detection end to the reactor outlet after daily opening. The measured values were recorded after stabilization. For the reactors under non-aerobic conditions, the same procedure cannot be applied since this would cause a larger amount of gases to be released from the reactor, thus affecting the accuracy of gas analysis for the remaining test (storage) period. Thus, TVOC concentrations associated with the non-aerobic reactors were measured at the end of the test period.

Data Analysis
In this study, the emissions of gases are expressed as emission factors, in units of [gram gas species per kilogram dry matter DM]. Emission factor is a cumulative parameter. The measured gas concentrations in vol% were converted to emission factors by using the N 2 balance method [13], assuming that nitrogen was not consumed during the test period. At constant temperature and pressure, the emission factor f in [g/kg DM] is related to volumetric gas concentration C i as follows: Energies 2020, 13, 128

of 14
where I is the gas species; P s is absolute pressure of gas in the container (Pa); C i is volumetric concentration of a particular gas (m 3 of gas species per m 3 of gas going through the GC); V is gas volume in the reactor (m 3 ); M wt is gas molecular weight (g/mol); C n0 is initial concentration of nitrogen (%); C nt is the concentration of nitrogen at time t (%); R is universal gas constant (8.31 J/(mol·K)); T is temperature (K); and m is the total mass of materials in the container (kg). The gas volume is the numeric difference between the volume of container and volume of chips. Change in pressure during the test period was assumed to be minimal in general.

Non-Aerobic Condition
The gas emissions from WRC chips with initial moisture content of 35% under non-aerobic condition are shown in Figure 3.
Energies 2020, 13, x FOR PEER REVIEW 5 of 14 where I is the gas species; Ps is absolute pressure of gas in the container (Pa); Ci is volumetric concentration of a particular gas (m 3 of gas species per m 3 of gas going through the GC); V is gas volume in the reactor (m 3 ); Mwt is gas molecular weight (g/mol); Cn0 is initial concentration of nitrogen (%); Cnt is the concentration of nitrogen at time t (%); R is universal gas constant (8.31 J/(mol·K)); T is temperature (K); and m is the total mass of materials in the container (kg). The gas volume is the numeric difference between the volume of container and volume of chips. Change in pressure during the test period was assumed to be minimal in general.

Non-Aerobic Condition
The gas emissions from WRC chips with initial moisture content of 35% under non-aerobic condition are shown in Figure 3.  The CO 2 emission factor (f CO2 ) increased gradually with respect to storage time for all temperatures due to its accumulation in the airtight reactors. Evidently, f CO2 was observed to be the highest at 20 • C, Energies 2020, 13, 128 6 of 14 followed by 35 • C and 5 • C, while f CO2 was the lowest under 50 • C. f CO2 ranged from 0.29 to 1.9 g/kg DM under all temperatures at the end of storage. By comparison, these values were found to be 20-30% lower than those obtained for the chips having higher moisture contents under the same temperatures.
CO was only observed to emit from the reactors under higher temperatures of 35 • C and 50 • C, and it was positively related to the temperature (Figure 3). The CO emission factor (f CO ) was observed to be two orders of magnitude less than the CO 2 emission factor. For the chips with initial moisture content of 35%, f CO increased gradually with time to 0.0012 g/kg DM and 0.02 g/kg DM at the end of the storage at 35 • C and 50 • C, respectively. The trends of the O 2 profiles were exactly opposite to the CO 2 profiles for all temperatures. The oxygen content decreased linearly as the produced CO 2 and CO increased. The lowest oxygen concentration was observed at 20 • C, which was in accordance with the highest CO 2 emission at this temperature.
The gas emission trends were similar to those for initial moisture contents of 20% and 50%. The highest CO 2 emission occurred at 20 • C. The CO emission was higher for WRC chips having higher initial moisture content at the same temperature. Biomass with higher moisture content may generate higher emissions of CO 2 due to biological process or chemical reaction process. By comparison, Figure 4 illustrates the CO 2 emissions from WRC chips with different initial moisture contents under 20 • C, 35 • C and 50 • C. It can be seen that CO 2 emission from the reactors was higher for WRC chips having higher initial moisture content. In Figure 4a, CO 2 was generated rapidly from reactors with the chips of 50% moisture content when the O 2 content was still high in the beginning of the storage test. As oxygen concentration in the reactor was eventually depleted to zero, f CO2 reached and was maintained at a plateau of 2.9 g/kg DM after 30 days. In contrast, f CO2 increased gradually for WRC chips with 35% and 20% moisture contents, attaining 1.9 g/kg DM and 0.42 g/kg DM respectively after two-month storage. In terms of the effect of storage temperature, CO 2 emissions from WRC chips stored at 20 • C were found to be the highest, followed by 35 • C storage. In contrast, when storage temperature is 50 • C, f CO2 values were smaller, ranging from 0.08 g/kg DM to 0.43 g/kg DM at the end of the test for the WRC chips at different moisture contents.
Energies 2020, 13, x FOR PEER REVIEW 6 of 14 The CO2 emission factor (fCO2) increased gradually with respect to storage time for all temperatures due to its accumulation in the airtight reactors. Evidently, fCO2 was observed to be the highest at 20 °C, followed by 35 °C and 5 °C, while fCO2 was the lowest under 50 °C. fCO2 ranged from 0.29 to 1.9 g/kg DM under all temperatures at the end of storage. By comparison, these values were found to be 20-30% lower than those obtained for the chips having higher moisture contents under the same temperatures.
CO was only observed to emit from the reactors under higher temperatures of 35 °C and 50 °C, and it was positively related to the temperature (Figure 3). The CO emission factor (fCO) was observed to be two orders of magnitude less than the CO2 emission factor. For the chips with initial moisture content of 35%, fCO increased gradually with time to 0.0012 g/kg DM and 0.02 g/kg DM at the end of the storage at 35 °C and 50 °C, respectively. The trends of the O2 profiles were exactly opposite to the CO2 profiles for all temperatures. The oxygen content decreased linearly as the produced CO2 and CO increased. The lowest oxygen concentration was observed at 20 °C, which was in accordance with the highest CO2 emission at this temperature.
The gas emission trends were similar to those for initial moisture contents of 20% and 50%. The highest CO2 emission occurred at 20 °C. The CO emission was higher for WRC chips having higher initial moisture content at the same temperature. Biomass with higher moisture content may generate higher emissions of CO2 due to biological process or chemical reaction process. By comparison, Figure 4 illustrates the CO2 emissions from WRC chips with different initial moisture contents under 20 °C, 35 °C and 50 °C. It can be seen that CO2 emission from the reactors was higher for WRC chips having higher initial moisture content. In Figure 4a, CO2 was generated rapidly from reactors with the chips of 50% moisture content when the O2 content was still high in the beginning of the storage test. As oxygen concentration in the reactor was eventually depleted to zero, fCO2 reached and was maintained at a plateau of 2.9 g/kg DM after 30 days. In contrast, fCO2 increased gradually for WRC chips with 35% and 20% moisture contents, attaining 1.9 g/kg DM and 0.42 g/kg DM respectively after two-month storage. In terms of the effect of storage temperature, CO2 emissions from WRC chips stored at 20 °C were found to be the highest, followed by 35 °C storage. In contrast, when storage temperature is 50 °C, fCO2 values were smaller, ranging from 0.08 g/kg DM to 0.43 g/kg DM at the end of the test for the WRC chips at different moisture contents. The cumulative CO2 emissions from WRC chips with different moisture contents at the end of two-month storage are listed in Table 1. Under non-aerobic condition, the values of fCO2 from WRC chips of 20% moisture content were found to be 80-85% lower than those from higher moisture content of 50% under the same temperatures. It confirmed that moisture content could promote degradation of materials during storage process. The cumulative CO 2 emissions from WRC chips with different moisture contents at the end of two-month storage are listed in Table 1. Under non-aerobic condition, the values of f CO2 from WRC chips of 20% moisture content were found to be 80-85% lower than those from higher moisture content of 50% under the same temperatures. It confirmed that moisture content could promote degradation of materials during storage process.

Aerobic Condition
Results of cumulative gas emissions obtained from the reactors with 35% moisture content WRC chips under aerobic condition are presented in Figure 5.

. Aerobic Condition
Results of cumulative gas emissions obtained from the reactors with 35% moisture content WRC chips under aerobic condition are presented in Figure 5.  With regard to temperature, the emissions of CO 2 and CO displayed similar trends to those from the non-aerobic reactors. After the entire storage period, the total CO 2 emissions from the aerobic reactors were greater. At 20 • C, the maximum f CO2 was almost 3 g/kg DM, which was higher than 1.9 g/kg DM for the non-aerobic reactor. For other temperatures (5 • C, 35 • C and 50 • C), the total CO 2 emissions ranged from 0.76 g/kg DM to 1.9 g/kg DM. Again, the values were greater for the aerobic reactors as compared to the non-aerobic reactors. This may be attributed to the important effect of oxygen on CO 2 generation. As a result of daily pumping of air into the aerobic reactors, the oxygen content remained high close to 20% effectively during the entire storage period.
CO emissions were generated from the reactors at 35 • C and 50 • C ( Figure 5). The cumulative emissions of CO were much lower than the CO 2 emissions. Furthermore, the cumulative CO emissions from the aerobic reactors had a similar range as those from the non-aerobic reactors. At the end of storage, the total CO emissions from WRC chips with 35% moisture content ranged from 8 × 10 −4 to 0.02 g/kg DM. As the air was pumped into the aerobic reactors, the moisture content of the materials decreased with time. It indicated that moisture content is a crucial factor affecting the CO emissions. With lower moisture content, the emission rate of CO decreased.
Cumulative CO 2 emissions from WRC chips with different moisture contents and for the three storage temperatures are illustrated in Figure 6. Emissions from the chips with different moisture contents had the same trend under different temperatures. At lower moisture content, f CO2 was lower under all temperatures. The lower moisture WRC chips also emitted a small amount of CO. As for the effect of storage temperature, the CO 2 emission factors at 20 • C were much higher than those from 35 • C and 50 • C. With regard to temperature, the emissions of CO2 and CO displayed similar trends to those from the non-aerobic reactors. After the entire storage period, the total CO2 emissions from the aerobic reactors were greater. At 20 °C, the maximum fCO2 was almost 3 g/kg DM, which was higher than 1.9 g/kg DM for the non-aerobic reactor. For other temperatures (5 °C, 35 °C and 50 °C), the total CO2 emissions ranged from 0.76 g/kg DM to 1.9 g/kg DM. Again, the values were greater for the aerobic reactors as compared to the non-aerobic reactors. This may be attributed to the important effect of oxygen on CO2 generation. As a result of daily pumping of air into the aerobic reactors, the oxygen content remained high close to 20% effectively during the entire storage period.
CO emissions were generated from the reactors at 35 °C and 50 °C ( Figure 5). The cumulative emissions of CO were much lower than the CO2 emissions. Furthermore, the cumulative CO emissions from the aerobic reactors had a similar range as those from the non-aerobic reactors. At the end of storage, the total CO emissions from WRC chips with 35% moisture content ranged from 8 × 10 −4 to 0.02 g/kg DM. As the air was pumped into the aerobic reactors, the moisture content of the materials decreased with time. It indicated that moisture content is a crucial factor affecting the CO emissions. With lower moisture content, the emission rate of CO decreased.
Cumulative CO2 emissions from WRC chips with different moisture contents and for the three storage temperatures are illustrated in Figure 6. Emissions from the chips with different moisture contents had the same trend under different temperatures. At lower moisture content, fCO2 was lower under all temperatures. The lower moisture WRC chips also emitted a small amount of CO. As for the effect of storage temperature, the CO2 emission factors at 20 °C were much higher than those from 35 °C and 50 °C. By comparing Figures 4 and 6, the total CO2 emissions generated were higher from the aerobic reactors after the storage test was completed. In the beginning of the storage test, fCO2 were similar for the non-aerobic and the aerobic reactors, but their differences became gradually larger afterwards. Over the storage period, the highest total CO2 emissions at 20 °C from the non-aerobic condition and the aerobic condition were 2.9 and 6.9 g/kg DM, respectively. The CO emission factor, fCO, was similar for both non-aerobic and aerobic conditions. Again, the difference between the CO2 emissions from non-aerobic and aerobic conditions could be attributed to the oxygen content in the reactors. As oxygen content dropped to zero, CO2 emission was maintained at a plateau value in the non-aerobic reactor. On the contrary, CO2 emissions increased continuously in the aerobic reactors amidst relatively high and constant oxygen level.

Emissions of NMVOCs
Gas emissions in terms of NMVOCs emissions were analyzed in this study. A range of chemical compounds were identified by GC/MS from the WRC chips under non-aerobic condition. The major NMVOCs emitted from WRC chips with different moisture contents were similar at all By comparing Figures 4 and 6, the total CO 2 emissions generated were higher from the aerobic reactors after the storage test was completed. In the beginning of the storage test, f CO2 were similar for the non-aerobic and the aerobic reactors, but their differences became gradually larger afterwards. Over the storage period, the highest total CO 2 emissions at 20 • C from the non-aerobic condition and the aerobic condition were 2.9 and 6.9 g/kg DM, respectively. The CO emission factor, f CO , was similar for both non-aerobic and aerobic conditions. Again, the difference between the CO 2 emissions from non-aerobic and aerobic conditions could be attributed to the oxygen content in the reactors. As oxygen content dropped to zero, CO 2 emission was maintained at a plateau value in the non-aerobic reactor. On the contrary, CO 2 emissions increased continuously in the aerobic reactors amidst relatively high and constant oxygen level.

Emissions of NMVOCs
Gas emissions in terms of NMVOCs emissions were analyzed in this study. A range of chemical compounds were identified by GC/MS from the WRC chips under non-aerobic condition. The major NMVOCs emitted from WRC chips with different moisture contents were similar at all temperatures, which included branched, straight and cyclic chain hydrocarbons, methanol, acetone, hexane, terpenes, ketones, aldehydes, benzene and its derivatives, acids and esters. Indole was also present in the Energies 2020, 13, 128 9 of 14 emission when 50% moisture content WRC chips were stored at a higher temperature of 50 • C. The cumulative concentration of total NMVOCs (TVOC) from non-aerobic reactors was measured at the end of two-month storage. Results showed that, at the same temperature, the TVOC concentrations had positive correlation with moisture content of the chips. In addition, TVOC concentrations increased with storage temperature for all three moisture contents. At the end of storage period, the highest measured TVOC concentrations were measured as 216 ppm, 482 ppm and 580 ppm (v/v) for WRC chips with moisture content of 20%, 35% and 50%, respectively.
Under aerobic condition, the detected components of NMVOCs were similar to those pertinent to non-aerobic condition. Figure 7a illustrates the profile of the cumulative TVOC concentrations from the aerobic reactors for WRC chips with 35% moisture content. It can be seen that the cumulative TVOC increased with time for all temperatures, though the daily increment of TVOC concentration decreased with time. In terms of temperature effect, NMVOC emissions from aerobic reactors and non-aerobic reactors exhibited similar trends, reaffirming that higher temperature could induce more release of VOCs as a result of organic matter degradation. The effect of moisture content on the cumulative concentrations of TVOC is depicted in Figure 7b. The trends of TVOC emissions were similar for the chips with different moisture contents. At the same temperature, higher moisture content led to the release of more NMVOCs. For instance, at 50 • C, at the end of the storage period, the summation of TVOC concentrations reached 356 ppm, 630 ppm and 811 ppm (v/v) for WRC chips with moisture contents of 20%, 35% and 50%, respectively.
Energies 2020, 13, x FOR PEER REVIEW 9 of 14 temperatures, which included branched, straight and cyclic chain hydrocarbons, methanol, acetone, hexane, terpenes, ketones, aldehydes, benzene and its derivatives, acids and esters. Indole was also present in the emission when 50% moisture content WRC chips were stored at a higher temperature of 50 °C. The cumulative concentration of total NMVOCs (TVOC) from non-aerobic reactors was measured at the end of two-month storage. Results showed that, at the same temperature, the TVOC concentrations had positive correlation with moisture content of the chips. In addition, TVOC concentrations increased with storage temperature for all three moisture contents. At the end of storage period, the highest measured TVOC concentrations were measured as 216 ppm, 482 ppm and 580 ppm (v/v) for WRC chips with moisture content of 20%, 35% and 50%, respectively. Under aerobic condition, the detected components of NMVOCs were similar to those pertinent to non-aerobic condition. Figure 7a illustrates the profile of the cumulative TVOC concentrations from the aerobic reactors for WRC chips with 35% moisture content. It can be seen that the cumulative TVOC increased with time for all temperatures, though the daily increment of TVOC concentration decreased with time. In terms of temperature effect, NMVOC emissions from aerobic reactors and non-aerobic reactors exhibited similar trends, reaffirming that higher temperature could induce more release of VOCs as a result of organic matter degradation. The effect of moisture content on the cumulative concentrations of TVOC is depicted in Figure 7b. The trends of TVOC emissions were similar for the chips with different moisture contents. At the same temperature, higher moisture content led to the release of more NMVOCs. For instance, at 50 °C, at the end of the storage period, the summation of TVOC concentrations reached 356 ppm, 630 ppm and 811 ppm (v/v) for WRC chips with moisture contents of 20%, 35% and 50%, respectively. The concentrations of TVOC emitted from WRC chips under aerobic condition were slightly higher than those under non-aerobic condition. For WRC chips with 35% moisture content and over two-month storage period, the cumulative TVOC concentrations were 13.5 ppm, 62.4 ppm, 258 ppm and 630 ppm (v/v) as temperature varied from 5 °C to 50 °C. These values are greater than those under non-aerobic condition, being 7.5 ppm, 20.7 ppm, 188 ppm and 482 ppm (v/v) at 5 °C, 20 °C, 35 °C, 50 °C, respectively.

Dry Matter Loss
Results pertinent to the dry matter loss of stored WRC chips after the test under non-aerobic condition and aerobic condition are summarized in Table 2. The concentrations of TVOC emitted from WRC chips under aerobic condition were slightly higher than those under non-aerobic condition. For WRC chips with 35% moisture content and over two-month storage period, the cumulative TVOC concentrations were 13.5 ppm, 62.4 ppm, 258 ppm and 630 ppm (v/v) as temperature varied from 5 • C to 50 • C. These values are greater than those under non-aerobic condition, being 7.5 ppm, 20.7 ppm, 188 ppm and 482 ppm (v/v) at 5 • C, 20 • C, 35 • C, 50 • C, respectively.

Dry Matter Loss
Results pertinent to the dry matter loss of stored WRC chips after the test under non-aerobic condition and aerobic condition are summarized in Table 2. The largest dry matter losses from WRC chips with 20%, 35% and 50% moisture contents were occurred at 20 • C, amounting to 0.02%, 0.15% and 0.25%, respectively. The loss in dry matter was higher from chips with high moisture content under the same temperature. It is evident that the losses of materials from all reactors were small with a rather narrow range of 0.01-0.25% after two-month storage.
As displayed in Table 2, the extent of dry matter loss was slightly higher from aerobic condition when comparing to those under non-aerobic condition. The greater dry matter losses under aerobic condition might result from the daily ventilation to replenish oxygen content in the reactors. It is worth noting that, for all the tests conducted in this study, the dry matter losses were low (<1%) when compared to those reported for other waste biomass materials in the literature.

Discussions
Biomass materials can be degraded both chemically and biologically. When chemical oxidation is dominant, the gas emissions increase with temperature according to the Arrhenius relationship for the temperature dependence of chemical reaction rates. Moreover, temperature is also one of the key factors that affect biological processes, which can be categorized with regard to three cardinal temperatures (minimum, optimum and maximum). When temperatures are lower or higher than the optimum, bacteria and fungi could be dormant or even killed. The highest f CO2 at 20 • C indicated that the generation of CO 2 from WRC chips was dominated by biological mechanism rather than chemical mechanism. CO is produced from chemical reaction of the materials and promoted by high temperature generally. The increase in the CO emission factor with temperature and oxygen availability in this study implied dominance by chemical mechanism.
The variation in the emission factors of WRC chips with different moisture contents suggested that the moisture content is an important factor affecting the gas emissions from biomass materials during storage. In terms of biological aspect, microbial activities have a vital influence on the process. Moisture content is known to be a key factor that governs biological reactions. It was found to have a significant effect on the size and activity of the microbial community as the overall patterns of respiration were strongly controlled by moisture content. The respiration rate of the microbes had a positive correlation with percentage humidity [34]. The microbial growth rate constant and the evolution of CO 2 increased linearly with increasing moisture content. And the microbial activity was inhibited under dry circumstance [35].
One previous study reported that the fresh forest residues with 60% moisture content (wet basis) emitted more than twice the amount of CO 2 as compared to that emitted from naturally dried forest residues with 40% moisture content after 6-month storage [29]. Other research also reported the importance of moisture content to the chemical reaction rate, as high moisture content favored the decomposition process and allowed rapid degradation circumstances [13]. Our results are in accord with these studies, that WRC chips having higher moisture content generated greater gas emissions.
For the purpose of better understanding, we compared our results with other woody materials (moisture content ranging from 3% to 50%, wet basis) in the literature. Yazdanpanah [36] measured the gas emissions from white pellets with moisture contents of 4%, 9%, 15%, 35% and 50% (wb) at temperatures of 25 • C, 40 • C and 60 • C for 9-week storage test period. The emission factors of CO 2 that ranged from 0.007 to 0.415 g/kg DM were positively correlated with temperature as well as moisture content. Our experimental results suggested that the CO 2 emissions from wood chips were an order of magnitude greater than those from the wood pellets. The differences in gas emissions may be due to the presence of live microorganisms in the wood chips, and the larger surface area of wood chips facilitate the microbial colonization. Furthermore, most of the microorganisms in wood pellets were killed during pelletization and pretreatment processes. Fresh wood chips could release more CO 2 than wood pellets in the beginning of storage, possibly as a result of biological reaction in addition to chemical oxidation. When oxygen was eventually depleted to zero, no more CO 2 was generated from biological process. Similar mechanism was also revealed in our earlier study on gas emissions from stored Douglas fir wood chips with moisture content of 50% (wet basis), whereby results indicated that CO 2 emissions increased from 3 g/kg DM to 18 g/kg DM as temperature rose from 5 • C to 50 • C [9]. The highest CO 2 emission from Douglas fir chips was about three times higher than that from the WRC chips. Besides, CO emissions from stored WRC chips were significantly lower when compared to the corresponding emissions from wood pellets.
The gas emissions of CO 2 and CO may have safety and health implications, depending on the duration of exposure. The threshold limit value-time weighted average (TLV-TWA) values for 8 h exposure to CO 2 and CO are 0.5 and 0.0025%, respectively. The threshold limit value-short term exposure limit (TLV-STEL) values for 15 min exposure to CO 2 and CO are 3% and 0.01%, respectively. The corresponding concentrations that are deemed "Immediately Dangerous to Life and Health" for CO 2 and CO are 4% and 0.12%, respectively [37]. Thus, the storage of biomass can be dangerous if it is not ventilated properly. In confined space, CO concentrations are several orders of magnitude higher than the occupational exposure limits. Serious accidents in Europe have been reported in the past decade, as the CO concentration was found to be high with a range of 14 ppm to 155 ppm (v/v) in the pellet storage bins [38].
As products of the degradation of woody materials during storage, many of the NMVOCs are highly flammable and may lead to the self-ignition during storage. The odorous NMVOCs (for instance aldehydes, ketones and terpenes) may have odor nuisance and even health problems, as they can be inhaled by people who live nearby or work close to the storage area. Some of the NMVOCs detected in this study were also reported in the literature. Monoterpenes were found to be the primary NMVOCs from fresh spruce and pine wood [4,39]. Wang et al. reported the emissions of aldehydes and ketones from stored wood pellets [25]. Aldehydes are the most abundant NMVOCs from pellet storage, in which hexanal accounts for 80% of emitted aldehydes. Other lab study also showed that the main NMVOCs including methanol, pentane, pentanal and hexanal were generated by the oxidation of stored pellets [38]. Temperature was in accord with the TVOC concentrations from stored WRC chips. It reaffirmed that higher temperature could induce more release of VOCs as a result of organic matter degradation. This phenomenon was also found from other biomass materials [25].
Dry matter loss is one of the major factors that attribute to the limitation of wood chips storage in large piles. Dry matter loss can affect the energy value of fuel, so that they have to be accounted for during the emission evaluations. The reported monthly dry matter loss from storage of fresh biomass chips ranged from 1-4% [26,28,29], while the maximum mass loss of lower than 1% was found in this study after two months of storage. By comparison with other woody materials reported in the literature, small degree of dry matter loss from this storage test reaffirms the decay-resistance characteristics of WRC chips as extractives like thujaplicins were present in the wood [3,9].
The utilization of forest chips has been predicted to increase in the future. Therefore, the safe storage of biomass is of great importance. The risks of storing wood chips include spontaneous heating and self-ignition, reduction of fuel quality, dry matter losses, human health and environmental risks from gas emissions [4]. To some extent, gas emissions during storage are much higher than those from the rest of the biomass supply chain. However, in practice, it may be difficult to monitor the gas emissions from wood chip piles. Results from this study can be used as a preliminary estimate to larger scale piles. In most cases, fresh wood chips contain high moisture when received, and the moisture content varies depending on sources and climate. Wood chips can be subject to drying process before storage, but it could have high energy consumption. Thus, it is desirable to grasp a good understanding of dry matter loss and the associated gas emissions.

Conclusions
This study concerns the measurements of gas emissions and dry matter loss from WRC chips during the storage process. Results suggested that the emission factors had positive correlation with moisture content of the chips. The highest CO 2 emission factors were found to be 2.9 g/kg DM and 6.9 g/kg DM from WRC chips with 50% moisture content (wet basis) under non-aerobic and aerobic conditions, respectively. A range of NMVOCs were detected from the storage tests, some of which were odorous that would induce safety and health problems. Moreover, the concentration of TVOCs was affected by moisture content of the chips. WRC chips with 50% moisture content had the highest TVOC concentration of 811 ppm (v/v) at 50 • C, whereas in the case of 35% moisture content, TVOC had the lowest value of 356 ppm (v/v). Dry matter loss at the end of the storage period was less than 1% for all tests, reaffirming the decay-resistance characteristics of WRC chips. Furthermore, CO 2 emission was dominated by biological mechanism whereas CO was generated by chemical oxidation mechanism under both non-aerobic and aerobic conditions.