Total organic compounds released from dehumidification drying of air-dried hardwood lumber.
Wood dry kilns have been identified by the Environmental Protection Agency as a potential source of significant volatile organic compound (VOC) and total organic compound (TOC) emissions. The VOC emissions from lumber drying are volatilized along with the water as both are evaporated from wet lumber. Very limited information is available on the quantity of VOC or TOC emissions from drying hardwoods and softwoods. The objective of this study was to quantify TOC amounts in commercial dehumidification kiln effluent that was used to finish drying air-dried hardwood lumber. Kiln effluent containing the TOCs for each charge was determined using a Shimadzu TOC-5000A emission measuring device both before and after an activated charcoal filter. The test results of the effluent collected before the activated charcoal filter were used to estimate the total amount of TOCs for each kiln charge. The effluent, after passing through activated charcoal, was analyzed to determine the effectiveness of the activated charcoal in removing TOCs and in reducing the pH of the effluent from the different kiln charges. The results indicated that a mixed charge of red oak and white oak lumber, dried from an initial moisture content of 21 percent, released the highest amount of TOCs of the six hardwood kiln charges in this study. The study also revealed that activated charcoal was effective in reducing TOCs from the effluent of the kiln charges. In general, the pH increased after charcoal filtration and TOCs decreased.
Volatile organic compounds (VOCs) are carbon compounds that take part in the formation of ozone, which is a common air pollutant and an environmental health hazard. Most recently, the focus on the emission of VOCs by regulatory agencies has started to shift to the wood products industry. VOCs from wood are thought to be derived mainly from the complex extractive compounds in wood. The amount of VOCs released per thousand board feet (MBF) of lumber may be relatively small, but as a cumulative annual amount for kiln-drying, the total may become quite significant. According to the U.S. Environmental Protection Agency (EPA), the total estimated VOCs released for the lumber and wood products industry is 41,423 short tons per year (EPA 1995).
Environmental regulations have had an economic impact on the forest products industry. Since the 1990 amendments to the Clean Air Act (CAA), the wood industry has been struggling to obtain useful information regarding emissions of VOCs. Title V of the CAA requires permits for significant sources of hydrocarbon emissions in accordance with state implementation plans (SIPs). Each permit must contain emission limits, record keeping and reporting procedures, a schedule for compliance, and a way for the state to ensure that there is compliance (Wu and Milota 1999). For certain segments of the wood industry, for example formaldehyde and finishing emissions, wood products companies are spending money and time conforming to regulations. The amount of money spent and the amount of testing required to achieve accurate results would be reduced if VOCs (and other emissions) could be accurately predicted (Shmulsky and Ingram 2000).
The main research focus on VOCs released from lumber has centered on softwood species. There are two primary reasons for this trend. First, softwood species comprise over 83 percent of the lumber dried in the United States (Rice 1995). Nearly all solid structural lumber products, as well as most engineered wood products (i.e., plywood and other structural panel products), used in construction are derived from softwood species. The second reason is related to the fact that conifers generally contain larger amounts of extractive material than hardwoods. Hardwood species, on the other hand, make up a smaller amount of the industrial lumber volume than softwood species because hardwoods are used primarily in aesthetic and appearance product applications (i.e., veneer overlays, cabinets, and furniture). This fact is directly related to the assumption that smaller total amounts of VOCs will be released during the annual drying of hardwoods than softwoods.
The process of drying hardwoods using a dehumidification (DH) kiln requires the condensation of the moisture along with the condensable water-soluble VOCs emitted from the wood during drying. The condensed effluent is usually discharged through a pipe. The only gaseous discharges from a DH kiln are through leaks around the door seals or, in some installations, small unpowered vents may purposely be installed to help equalize pressure between the inside of the kiln and the ambient conditions. Measuring the total organic compound (TOC) content in the DH kiln effluent should provide a reasonable estimate of the condensable VOCs in drying hardwood lumber.
Limited research has been reported on quantifying the amount of VOCs from hardwood lumber species. The main objective of this study was to quantify the amount of TOC emitted during DH drying of several different charges of air-dried hardwood lumber. Another objective was to evaluate the effectiveness of activated charcoal on reducing TOCs from the effluent of a DH kiln while drying different charges of air-dried hardwood lumber.
TOC data collection
The commercial DH kiln that was used for this study had a capacity of approximately 10 MBF. Data were collected on six charges of air-dried hardwood lumber. For each charge, the kiln operator loaded the air-dried lumber into the DH kiln and started the drying schedule. The average initial moisture content (MC) of each kiln charge ranged from 14.5 to 25 percent (Beakler 2002). As moisture was removed from the wood, it was condensed along with a portion of the VOCs by the DH units inside the kiln and discharged through a pipe that served as the kiln effluent drain.
Sampling of the DH kiln water effluent began the first day of each DH kiln charge. The kiln operator used an electronic scale, timer, and bucket. A bucket was used to catch the effluent that was leaving the kiln through the polyvinyl-chloride (PVC) condensate drain. The bucket collected effluent at a selected time for each day of the kiln schedule. A timer was set when the operator started collecting the effluent, and after 30 minutes or 1 hour of collection, the bucket was removed and weighed. The effluent weight information was recorded and a sampling jar (approximately 270 mL) was filled with the condensate at the condensate drain and placed in a refrigerator. This procedure was executed each day during the DH kiln run. The kiln operator recorded the time and date of sampling, kiln conditions (temperature and relative humidity), pounds per unit time of condensate collected, sample jar number, species, thickness of lumber, volume of lumber, initial and final estimated MCs, and any remarks concerning kiln conditions. The daily weight of the effluent in the sample bucket was recorded and used to estimate the total amount of effluent condensed in the DH kiln for each day. The kiln operator also used the daily weight of the effluent to estimate the drying rate and the average MC of the kiln charge. The estimated daily pounds of water condensed was based on the measured daily pounds of water per unit time. The measured daily pounds of water per unit time was assumed to be representative of a 24-hour period and was also used with the daily TOC samples to estimate the daily TOC amounts in the effluent. The average initial and the final average MCs of the DH kiln charge were measured by the kiln operator.
Effluent discharge from the DH kiln passed through a filter bucket containing activated charcoal. After the kiln operator obtained the effluent sample (before the activated charcoal), the bucket of activated charcoal was removed from its storage area. A sample of water was taken as the discharge water leaked through the perforations in the bottom of the activated charcoal bucket. As in the first sample before the activated charcoal bucket, approximately 270 mL of effluent was collected, numbered, and placed in a refrigerator.
After each drying schedule was completed, all of the effluent samples were removed from the refrigerator and placed in a cooler for transport to the laboratory for pH and TOC content determination. The testing instrument used on the effluent samples was the Shimadzu TOC-5000A. The pH was first determined for each sample, after which the samples were poured into small test tubes and hydrochloric acid was added to each test tube to lower the sample pH to 2. The TOC machine was purged with purified air to remove any inorganics, and dry combustion was completed on each sample at 680[degrees]C over a platinum catalyst.
A conversion system was developed to convert the daily results of the TOC analysis from mg/L to pounds of TOC. The measured daily condensed effluent and the daily TOC data were utilized for these calculations (Beakler 2002). The conversion started with mg/L of TOC and resulted in pounds of TOC/L per day (mg/L X g/1000 mg X 0.0022 lb./g = lb. TOC/L). The second conversion step started with estimated pounds of kiln discharge water per day (estimated from the measured effluent amount each day) and resulted in gallons of kiln discharge water per day (lb. of discharge water X 1 gal. water/8.3378 lb. water = gal. of kiln discharge water). The third conversion step started with gallons of kiln discharge water per day and resulted in liters of kiln discharge water per day (gal. of kiln discharge water X 3.78541 L/gal. = L of kiln discharge water). The last conversion step uses liters to yield pounds of TOC per day (L of kiln discharge water X lb. of TOC/L = lb. of TOC). Each day during the DH kiln run the TOC in the discharge effluent was determined. All of the daily TOC estimations were summed to yield an estimated total TOC loss for the entire DH kiln run. This number would be the TOC discharge for the amount of BF being dried in the DH kiln.
There were six charges of 4/4 hardwood lumber analyzed in this study. All of the lumber utilized in this research project was air-dried in an outdoor environment and placed in the DH kiln at different starting MCs. In addition, the drying was stopped at slightly different final MCs. Table 1 lists the species composition and total volume in BF of the charges analyzed. Since the composition of each charge was different in either total charge BF or range in MC reduction (initial MC--final MC), the data analyses were based on either a MBF or total range in MC (initial--final) for each DH kiln charge.
The kiln operator determined an estimate of the MC of the lumber each day during the drying cycle. The MC of the air-dried lumber at the start of DH drying was obtained using a resistance-type moisture meter before the lumber was stacked in the kiln. An average MC was determined across the lumber stacks and used as the initial MC. Each day during the kiln run, the weight of the effluent (exiting the kiln as water) per half-hour or hour was recorded on the kiln emission data sheet. The amount of kiln effluent per unit time was used to adjust the DH kiln atmosphere and determine the end of drying. The kiln door was only opened at the start and finish of drying. The estimated total pounds of water for each day were determined and summed over the complete drying cycle to obtain the total pounds of water per each drying charge.
One-way analysis of variance (ANOVA) yielded Tukey multiple comparison tests that were used to determine statistical significance between means. All Tukey tests were carried out at an alpha level of 0.05. F-values and p-values were also obtained from one-way ANOVA and used to test hypotheses in the analysis of equal means. Paired t-tests were also used in the activated charcoal analysis.
Results and discussion
Table 2 lists an estimate of the total amount of TOC released (in lb.) for each kiln charge. An unknown portion of the VOCs released during DH drying may not have been condensed along with the evaporated moisture. An estimate of this unknown portion was not attempted. Calculations yielded pounds of TOC per percentage change in MC for that charge (Table 3). The volume in MBF was determined for each charge and pounds of TOC per percentage change in MC was divided by MBF. This calculation yielded pounds of TOC per percentage change in MC per MBF of lumber (Table 2). All estimates (Table 2) were on an equal basis level for statistical comparison.
The first mixed hardwood charge released the second highest amount of TOCs of all the charges. The second mixed hardwood charge contained approximately 74 percent soft maple, which likely had an effect on TOC release. Yellow-poplar comprised 70 percent of the third mixed hardwood charge, which also may have influenced total TOC release.
For the poplar charge, 0.00433 pounds of TOC were released per percentage change in MC (12% MC range from initial to final MC). Total organic compounds per percentage MC change per MBF of yellow-poplar lumber was 0.00046 pounds.
The first charge of red and white oak released 4.5 times more TOCs than the second highest amount. The large difference between the TOC released for the two red and white oak charges could be that the first red and white oak charge had a 14.5 percent MC range from initial to final MC and the second charge of red and white oak had an 8 percent MC range.
A one-way ANOVA was used to analyze the average total pounds of TOC lost per percentage change in MC for the six DH kiln charges. A Tukey multiple comparison test on the six average values determined which of the averages were statistically similar. Results of the Tukey test ([alpha] = .05) revealed that the total average TOC lost per percentage change in MC for the first red and white oak charge was significantly different from all the other kiln charges. The last charge of red and white oak was also significantly different from all other charges except for the first charge of mixed hardwood species. All of the other charges were statistically similar according to the Tukey multiple comparison test. An extremely high F-value, and a low p-value indicated that all the mean values were not equal, reinforcing the results of the Tukey multiple comparison test.
Hardwood extractives contain low molecular weight organic compounds that can be easily volatized with the evaporation of water, as well as higher molecular weight compounds that can be broken down into low weight components at elevated temperatures. Approximately 2 to 7 percent extractive content occurs in the stemwood of most hardwood species (Koch 1985). Heartwood and bark of many wood species contains a higher extractive content than the sapwood, due to high content of phenolic extractives (Sjostrom 1993).
There are several components in hardwood extractives that could possibly be released as VOCs, but there has been limited research done to determine these specific compounds. Tannins are the water-soluble polyphenols found mainly in hardwoods and they range in molecular weight from 500 to 3000 (Koch 1985). There is a possibility that some of the low molecular weight tannins can be driven off the lumber as VOCs. Tannins that can be readily hydrolyzed by acids or enzymes are considered hydrolyzable tannins (Koch 1985). Gallotannins and ellagitannins are types of hydrolyzable tannins that are present in measurable amounts in hardwoods. Condensed tannins are the last type of tannins found in hardwoods and are flavonoid polymers that have the highest molecular weight of any type of tannin. Due to their high molecular weight, extremely small amounts escape as VOCs, unless they are degraded and volatilized at elevated temperatures. Tannins are the most prevalent extractive group in the hardwood family.
There are several other types of extractive compounds that have potential to be released as VOCs. Sterols, pinoresinol, syringaresinol, catechin, and lignans are a few of the minor extractive species that can be identified from hardwoods (Koch 1985). Knotts et al. (1995) found that DH drying of red oak produced identifiable amounts of butyrolactone, butanoic acid, 2-chloro-butanoic acid, and bis(2-ethylhexyl)phthalate in kiln condensate samples. Other organic compounds such as ellagic acid, gallic acid, and hamamelitannin may also be present in red oak kiln condensate (Rowe and Connor 1979). There are large varieties of hardwood extractives that usually occur in small amounts. Any of these extractive species, depending on molecular weight, can be volatilized.
Knotts et al. (1995) indicated a low of 238 parts per million TOC and a high of 1286 parts per million TOC in red oak DH kiln effluent. The study by Knotts et al. (1995) included high initial MCs. The results of the first charge of red and white oak exhibited a range of 180.6 mg/L TOC (low) to 355.8 mg/L TOC (high). The average TOC (mg/L) for the first charge of red and white oak falls in the lower end of the range indicated in the study by Knotts et al. (1995). The relatively low TOC results for the lumber used in this study are not surprising since the lumber was air-dried prior to being dried in the commercial DH kiln.
Charcoal filtration effects on TOC effluent
The effluent from the DH kiln was treated with activated charcoal before it was discharged into an underground limestone bed. Effluent samples were taken before the charcoal filter and after the charcoal filter so the effectiveness of the activated charcoal could be determined. Activated charcoal was replaced before charges 2 and 4. An increased percentage of TOC removal was measured using new activated charcoal filters. Table 3 lists the average amount of TOC mg/L both before and after activated charcoal treatment and the percentage of TOC removed by the charcoal. A one-way ANOVA analysis ([alpha] = 0.05) on each of the six kiln charges used the data before and after the charcoal treatment. The analysis of variance results indicated that the means before and after treatment for each charge were significantly different. The statistical analysis also computed Tukey intervals for the one-way ANOVA. The Tukey analyses indicated that all of the means were statistically different. These statistical tests showed that the activated charcoal was effective in lowering the TOC average. However, even though the charcoal was effective in all of the charges, the data indicated that the charcoal filter worked at different levels of effectiveness for the different kiln charges. High amounts of TOC after passing through the charcoal may be related to a degree of charcoal saturation or inability of this type of activated charcoal to remove certain portions of the TOC contents produced by certain hardwood species in the effluent.
Large-diameter activated charcoal particles, like the ones used in this study (sieve tests on samples of the activated charcoal yielded an average of 47.4% by weight to be 0.0394-in. particles), may have less available surface area for the activated carbon to react with TOC compared to smaller particles. Thus, there is a possibility that TOC test results for all kiln charges could be improved if smaller activated charcoal particles were used. In addition, the type of activated charcoal that was used in this study was designed for filtration of residential water sources and may not be optimum for removing organic compounds released during the drying of hardwood lumber.
Tests for the pH of each effluent sample both before and after charcoal filtration were conducted. Results of pH analysis showed that all of the effluent samples that were tested before and after the charcoal filter were acidic (Table 4). The first red and white oak charge had the lowest average pH. Solliday et. al (1999) reported that DH drying of 4/4 red oak lumber produced acetic acid in the kiln condensate. The average pH after the charcoal filter was less acidic for all charges, indicating that some of the acidic compounds were removed from the effluent by the activated charcoal.
For all charges, paired t-tests indicated that the average pH after the charcoal treatment was statistically different from the pH before the charcoal treatment. A one-way ANOVA was executed for the average differences in pH between the six kiln runs. A Tukey multiple comparison test ([alpha] = .05) indicated that all of the average differences in pH were statistically similar with the exception of the poplar charge. The difference in poplar pH was statistically different from all of the other charges. The kiln operator changed the activated charcoal before the poplar kiln charge was run, and this may have influenced the pH data. Activated charcoal was more effective in removal of acidic components of the effluent for the poplar charge. These statistical tests show that activated charcoal was effective in raising the pH of the kiln effluent of the six kiln charges tested.
The activated charcoal was effective in the removal of a portion of the acidic components from the kiln effluent, but when the effluent left the charcoal filter it was still acidic. The effluent at this commercial operation, after it passed through the activated charcoal filter, passed through an underground pit filled with limestone. Limestone is one of the most widely used agents to neutralize acidic solutions. Samples were not obtained after the limestone treatment. Solliday et. al. (1999) reported that acetic acid presents no serious environmental hazards after the low pH (from red oak DH kiln effluent) is neutralized with soda ash. The effectiveness of the limestone in neutralizing the effluent in this study is unknown.
The first three kiln charges all contained mixed hardwood species. The total TOC releases were estimated to be 0.2938, 0.0355, and 0.0448 pounds, respectively. Since these charges contained anywhere from 5 to 10 different species, it is impossible to quantify TOC amounts for each species. The fourth kiln charge consisted of 9,500 BF of yellow-poplar. The poplar charge released an estimated total of 0.0519 pounds of TOC in the kiln effluent. The last two charges contained both red and white oak lumber. The first charge of red and white oak lumber exhibited the highest estimated TOC release at 1.3748 pounds, while the second charge had the third highest estimated TOC release at 0.2648 pounds. The oak charges rank as the top two, respectively, in pounds of TOC per percentage change in MC and pounds of TOC per percentage change in MC per MBF of lumber.
Activated charcoal was effective in lowering the TOC amounts from the kiln effluent of each charge before it was released to the environment. The activated charcoal was more effective in removing TOC from the charges that released lower amounts of total TOC. A degree of saturation or inability of the activated charcoal to remove some components of the TOC in the effluent stream may have caused the activated charcoal to lose some of its effectiveness. The type of charcoal used at the commercial operation was designed for a residential water filtration system, and may not be the best type of charcoal for the removal of organic compounds released during the drying of wood.
Table 1.--Species composition of the DH kiln charges. 1 2 3 4 5 6 (BF) Ash 0 0 500 0 0 0 Beech 2,400 1,000 0 0 0 0 Black cherry 1,300 70 1,000 0 0 0 Black walnut 0 0 500 0 0 0 Elm (American and slippery) 250 0 0 0 0 0 Hard maple (sugar maple) 700 0 0 0 0 0 Hickory (mockernut, shagbark, and shellbark) 500 0 0 0 0 0 Red oak (scarlet, northern red, black, and pine) 700 0 0 0 8,000 5,850 Soft maple (red maple) 1,900 8,000 1,000 0 0 0 Sweet birch 900 600 0 0 0 0 Sweet gum 0 70 0 0 0 0 White oak 700 100 0 0 1,500 3,200 Yellow-poplar 700 1,000 7,000 9,500 0 0 Total 9,500 10,840 10,000 9,500 9,500 9,050 Table 2.--Estimated TOC for each DH kiln charge. Pounds of TOC/ Total range in MC Pounds of TOC/ % change in MC/ Charge Total TOC (initial--final) % change in MC MBF of lumber (lb.) (%) (lb.) 1 0.2938 18.0 0.01632 0.00172 2 0.0355 10.5 0.00338 0.00031 3 0.0448 6.0 0.00747 0.00075 4 0.0519 12.0 0.00433 0.00046 5 1.3748 14.5 0.09481 0.00998 6 0.2648 8.0 0.03310 0.00366 Table 3.--Percentage of TOC removal by activated charcoal for each DH kiln charge. DH kiln charge 1 2 3 4 Avg. TOC (before charcoal) for total run (mg/L) 106.1620 26.6000 23.3193 37.2833 Avg. TOC (after charcoal) for total run (mg/L) 78.9875 0.9689 20.7583 5.8833 TOC removed by charcoal (%) 25.60 96.36 10.98 84.22 DH kiln charge 5 6 Avg. TOC (before charcoal) for total run (mg/L) 311.3970 121.5600 Avg. TOC (after charcoal) for total run (mg/L) 213.6230 38.7700 TOC removed by charcoal (%) 31.40 68.11 Table 4.--pH data before and after activated charcoal for each DH kiln run. Avg. pH Range of pH Charge Before charcoal After charcoal Before charcoal After charcoal 1 3.92 4.23 3.77 to 4.12 4.16 to 4.27 2 4.52 5.58 4.32 to 4.71 5.37 to 5.68 3 4.10 4.01 3.84 to 4.60 3.90 to 4.32 4 4.21 6.40 4.15 to 4.31 6.24 to 6.52 5 3.50 4.21 3.39 to 4.04 3.72 to 6.54 6 3.54 4.22 3.49 to 3.58 3.92 to 5.58
Beakler, B. 2002. Quantification of volatile organic compounds released from dehumidification drying of air-dried hardwood lumber. MS thesis. School of Forest Resources, Pennsylvania State Univ., University Park, PA. 86 pp.
Environmental Protection Agency (EPA). 1995. EPA Office of Compliance sector notebook project profile of the lumber and wood products industry. Section IV.C. Exhibit 20. U.S. EPA, Office of Compliance, Washington, DC. 56 pp.
Knotts, D.F., J.P. Armstrong, and D.J. Gardner. 1995. Assessment of water quality from dehumidification kilns. In: Proc. No. 7301. Forest Prod. Soc., Madison, WI. pp. 29-34.
Koch, P. 1985. Utilization of Hardwood Growing on Southern Pine Sites. Agri. Handb. 605. USDA Forest Serv., Washington, DC. 3710 pp.
Rice, R.W. 1995. Assessing human health and environmental effects related to drying wood. In: Proc. No. 7301. Forest Prod. Soc., Madison, WI. pp. 14-16.
Rowe, J.W. and A. Connor. 1979. Extractives in Eastern hardwoods--A review. Gen. Tech. Rept. FPL-18. USDA Forest Serv., Forest Prod. Lab., Madison, WI. 23 pp.
Shmulsky, R. and L.L. Ingram, Jr. 2000. Empirical predictions of VOC emissions from drying southern yellow pine lumber. Forest Prod. J. 50(6):61-63.
Sjostrom, E. 1993. Wood Chemistry Fundamentals and Applications. Academic Press, San Diego, CA. 293 pp.
Solliday, D.S., J.P. Armstrong, and B.E. Dawson-Andoh. 1999. Dehumidification drying of red oak. Part 1. Chemical characterization of volatile organic compounds. Forest Prod. J. 49(7/8):21-23.
Wu, J. and M.R. Milota. 1999. Effect of temperature and humidity on total hydrocarbon emissions from Douglas-fir lumber. Forest Prod. J. 49(6):52-60.
Brian W. Beakler*
Paul R. Blankenhorn*
Lee R. Stover
Charles D. Ray*
The authors are, respectively, Research Assistant, Professor of Wood Technology, Senior Research Assistant, and Assistant Professor of Wood Products Operations, School of Forest Resources, Pennsylvania State Univ., University Park, PA 16802. This paper was received for publication in July 2003. Article No. 9709.
*Forest Products Society Member.
|Printer friendly Cite/link Email Feedback|
|Author:||Beakler, Brian W.; Blankenhorn, Paul R.; Stover, Lee R.; Ray, Charles D.|
|Publication:||Forest Products Journal|
|Date:||Feb 1, 2005|
|Previous Article:||Stress distribution in structural composite lumber under torsion.|
|Next Article:||Environmental concentrations of copper, chromium, and arsenic released from a chromated-copper-arsenate-(CCA-C-) treated wetland boardwalk.|