with an electrical readout. The temperature of 

 the stack gases was measured using a chromel 

 alumel thermopile. Data from the pitostatic 

 probe and the thermopile were used to give 

 stack flow rate as a function of time. This in- 

 formation, along with the sampling rate, al- 

 lowed us to determine what percentage of the 

 total effluent from the fire was being sub- 

 sampled at any given time. 



Filter papers were dried and preweighed 

 before a test fire. A filter paper was placed in 

 the particulate sampler; upon ignition, the 

 sampler was turned on and flow rate and time 

 were recorded. These readings were recorded 

 at intervals corresponding to a drop in flow 

 rate of 1 ft. 3 min." 1 or 5 minutes, whichever 

 occurred first. A given filter paper was 

 changed when the indicated flow rate was re- 

 duced to 5 ft. 3 min.- 1 (0.142 m. 3 min." 1 ). The 

 changes involved a standard 1-minute shut- 

 down time. This procedure was continued for 

 60 minutes. 



Contaminated filters were dried and 

 weighed to obtain particulate deposited dur- 

 ing the time the filter was in place. Sample 

 flow rate was corrected to standard condi- 

 tions and then corrected to experimental con- 

 ditions. The average flow rate and the total 

 flow between readings were then computed. 



The sums of the flows for each filter were di- 

 vided into the particulate accumulated on the 

 filter to give particulate density (mg.m." 3 ) as a 

 function of time. The densities between filters 

 were interpolated from the adjacent filter 

 data. The rate of particulate emitted and the 

 total particulate produced by the fire were 

 obtained by relating the data from the sam- 

 pler back to stack conditions. 



The approximate tar content of the partic- 

 ulate was determined by extraction of filter 

 paper subsamples using tetrahydrofuran. The 

 elemental phosphorus content of the particu- 

 late from the DAP treated samples was deter- 

 mined using acid digestion and colorimetry. 

 Water soluble sulfate present in the particu- 

 late was determined turbidimetrically using 

 Sulfaver, a product of Hach Chemical Com- 

 pany, Ames, Iowa. 7 



The data were statistically analyzed using 

 Program SCRAP (6.0.003) on an IBM 1620 

 computer. 



1 Norman A. Huey. Determination of sulfate in 

 atmospheric particulate: turbidimetric barium sulfate 

 method. P. 11-14, in: Selected Methods for the 

 Measurement of Air Pollutants. R. A. Raft, Ed. 

 USDHEW Public Health Serv., Cincinnati, Ohio. 

 1965. 



6 



