Historic, archived document
Do not assume content reflects current scientific knowledge, policies, or practices.
“a
‘wy ‘ .
rf
he
ne
roa
nee
oe MA A Ul
United fe! HAT Ate oe 0 7 gene CoP1°? Monoammonium | biahas Phosphate: Effect
Forest and Range Experiment Station sare” =~ QT Flammability of
Research Paper 313
om Excelsior and Pine | rt Needles 83 3072
MONTHLY ag MONTHLY A MONTH April ; ham FF ney 186.
Aylmer D. Blakely
RESEARCH SUMMARY
Treated ponderosa pine needle and aspen excelsior fuel beds were burned to quantify the fire-retarding capabilities of five samples of monoammonium phosphates. The samples were the same except for their manufacturing processes, which could cause chemical impurities that may decrease combustion-retarding effectiveness. Treat- ment levels were normalized by converting to phosphate equivalents when comparing effectiveness. To obtain about equal penetration, approximately 0.26 gal (1 liter) of a solution containing one of the chemicals was applied to each bed. Different chemical treatment levels were obtained by varying the solution concentrations. Solutions were sprayed onto the fuel beds from a fan-type nozzle, and after drying completely, the fuel was burned ina 5-mi/h (8-km/h) wind at 90° F (32.2° C) and 20 percent rela- tive humidity. Analysis of covariance and percent reduction in combustion rates were methods used to compare levels of effectiveness. Test results indicate no significant differences between the combustion-retarding abilities of the five monoammonium phosphates, and all proved to be as effective as standard diammonium phos- phate when compared at equivalent phosphorous applica- tion levels.
THE AUTHOR
AYLMER D. BLAKELY received his B.S. degree in forestry in 1960, and his M.S. degree in forestry in 1970 from the University of Montana. In 1967, he joined the Inter- mountain Station’s Northern Forest Fire Laboratory in Missoula, Mont., where he currently works in fire retardent chemical and delivery systems research in the Fire Control Technology Research Work Unit.
The use of trade, firm, or corporation names in this publication is for the information and convenience of the reader. Such use does not constitute an official endorsement or approval by the U.S. Department of Agriculture of any product or service to the exclusion of others that may be suitable.
Monoammonium Phosphate: Effect on Flammability of Excelsior and Pine Needles
Aylmer D. Blakely
INTRODUCTION
Forest fire retardants were first operationally delivered by air- craft in the mid-1950’s, and consisted mainly of chemicals that retained water and amassed thick layers on aerial fuels. The first chemical to be used extensively was sodium-calcium borate (commonly called ‘‘borate’’) that not only thickened the water, but had some fire-retarding properties when dry (Miller and Wilson 1957). Borate was toxic to plants, erosive on pumps, and costly because of the required mix ratio. Bentonite clay was introduced to overcome some of the problems caused by borate, but it only thickened the water and had no long-term retarding properties when dry (Phillips and Miller 1959). Borate and bentonite use was phased out when chemicals (called long- term retardants) were introduced that more effectively retard combustion, even when completely dry.
Experiments with long-term chemicals were performed by Truax (1939) to quantify the fire-retarding abilities of water solutions of several commonly used chemicals and chemical combinations, and to determine the feasibility of using them on wildfires. These were some of the same chemicals that had been used successfully for impregnating and fireproofing building materials since the early 1900’s. Studies by Truax and later by Tyner (1941) showed that diammonium and monoammonium phosphate water solutions were the most effective for retarding combustion. Other phosphate compounds have been tested, but the ammonium phosphates are the most chemically available for affecting pyrolysis. George and others (1977) reflect that phosphate compounds formed with Fe, Ca, or Mg usually are ineffective because of their high temperature requirements for decompositions and thus their unavailability in terms of altering pyrolysis and combustion reactions. Operation Firestop (1955a, 1955b) tested some phosphate chemicals along with borate, but because of the test methods and interpretation of results, _ borate was considered the superior fire-retarding chemical, and thus its use as the original aerial-delivery fire retardant.
Monoammonium phosphate (MAP) was dropped from a TBM airtanker onto forest fires in Georgia (Johansen 1959) with good results. Soon after, attempts were made to thicken the MAP with clays or gums (Johansen and Shimmel 1963) for better adherence to aerial fuels. Use of MAP was abandoned when Pyro, a liquid mixture of ammonium phosphate species, was introduced in the Southeast (Johansen and Crow 1965). Pyro was comparatively inexpensive and required only dilution with water and pumping into the airtanker. MAP was consid- ered unhandy because it required breaking bags open and mix- ing with water under agitation. Later developments of mixing equipment resolved most of the limiting mixing procedures.
In 1961 tests were performed (Hardy and others 1962) to compare the effectiveness of several different retardant formu- lations that were being suggested and introduced by firefighters and chemical companies. Among those chemicals were borate, algin-gel, diammonium phosphate (DAP) thickened with algin and pectin, and ammonium sulfate thickened with attapulgite clay. The tests showed that the sulfate- and phosphate-based materials were superior to borate and water thickeners, espe- cially after all the water had evaporated. Fire-Trol® (formu- lated with ammonium sulfate and clay thickener) and Phos- Chek® (formulated with diammonium phosphate and gum thickener) brand fire retardants were first produced commer- cially about 1962, and formulations containing various dry chemical combinations have since been the principal fire retard- ants used in the United States for combating wildfires. Several studies have been conducted to better quantify the combustion- retarding effectiveness of sulfates and phosphates, and to iden- tify their basic fire-retarding mechanisms (George and Susott 1971; George and Blakely 1972; Browne and Tang 1963; and Eickner 1962). Other phosphate-based retardants have since been used extensively: Pyro, previously mentioned, and Fire- Trol 931, made of 10-34-0 ammonium polyphosphate (Wood 1970; George 1971; George and others 1977).
In recent years, many of the chemicals used to prepare retardant formulations—basic retardant chemicals as well as additives for coloring and corrosion inhibition—have increased severalfold in price and, in some cases, have become difficult or impossible to obtain. For example, for several years the DAP used in retardants was produced as a byproduct from the conversion of coal to coke. Phosphoric acid (H; PO,) was used to remove (scrub) ammonia bearing-off gases from manufactur- ing effluents. In this reaction, either mono- or diammonium phosphate is produced, but unfortunately, the cost of using these products has become prohibitive. MAP and DAP are also manufactured by bubbling ammonia gas through H;PO,. Much of the high cost of combining ammonia and acid is in the costs of ammonia or nitrogen (N); therefore, it is more economical to add only one ammonia (NH;) to each phosphate (PO,) to make NH,H,PO,(MAP). In other cases, retardant users are searching for chemicals that are less expensive because of their formulas and/or manufacturing processes, but which are still cost-effective.
This study quantified the fire-retarding effectiveness of monoammonium phosphate chemicals from different sources and compared their effectiveness to diammonium phosphate (the basic fire retardant chemical in some currently approved retardants). Tests were performed with five MAP samples manufactured by various companies and/or processes.
The samples were basically the same except for minor differ- ences in composition and manufacturing process. The chemicals and their apparent differences are:
M-MAP - Fisher Scientific, ACS grade, granular form.
Contains less than 0.03 percent impurities.
S-MAP - Technical grade, granular form. Produced from
* technical grade phosphoric acid (white acid process) that has been neutralized with anhy- drous ammonia.
D-MAP -— Technical grade, crystalline form. Technical grade phosphoric acid neutralized by ammonia- rich gases that are given off by burning coal to make coke.
T-MAP - Technical grade, crystalline form. Produced from technical grade phosphoric acid that has been neutralized with anhydrous ammonia. (T-MAP and S-MAP are manufactured by simi- lar processes, but by different companies.)
A-MAP -— Technical grade, crystalline form. Manufactured from wet-process phosphoric acid and then re- purified to produce a technical grade product with less than 0.1 percent of impurities.
The phosphate quantities were calculated as P.O; equivalents to aid effectiveness comparisons because all MAP or DAP for- mulations should be equally effective if the phosphate (P.O;) in each (the principal fire-retarding element) is present and avail- able in equal concentrations.
The study was an indirect way to determine if the manufac- turing process, manufacturer, or quality of fire retardant chem- ical causes any reduced availability of the phosphorus at the precise time or temperature during pyrolysis when the retardant would have the needed effect (George and Susott 1971). Com- parison of like quantities of different chemicals that are equally effective fire retardants will indicate how much added chemical is necessary to increase the total available phosphorus in a par- ticular chemical formula. All effectiveness data are compared to DAP because it has been used as a standard of comparison since 1970 (George and Blakely 1972).
STUDY METHODS Fuel Beds
Combustion-retarding effectiveness data were gathered by burning mat-type fuel beds treated with different retardant chemicals. Two fuels were used for the study—ponderosa pine needles and aspen excelsior. The pine needles, gathered locally, were cleaned of debris, grass, and sticks, and stored to dry. The excelsior was ordered in compact bales that were pulled apart and allowed to come to equilibrium moisture content under inside conditions. The two fuels were used to determine the fire-retarding effectiveness of chemicals on fuels with a high cellulose (42 percent) and a low crude fat (1 percent) content (excelsior), and fuels with a low cellulose (18 percent) and a high crude fat (10 percent) content (needles). These fuels were not necessarily used to duplicate natural wildfire conditions,
but because they (1) are relatively easy to obtain, (2) respond to temperature and humidity changes, (3) are similar in chemical content to many fuels encountered on wildland fires, and (4) provide predictably reproducible fires under controlled environ- mental and fuel moisture conditions.
Standard techniques were used (George and Blakely 1972) for constructing and treating the 8-ft- (2.44-m-) long, 3-inch-(7.62-cm-) deep, 18-inch- (45.72-cm-) wide fuel beds. Each pine needle bed contained 6 lb (2.72 kg) of fuel, and each excelsior bed contained 4 lb (1.81 kg) of fuel. The fuel moisture content (measured by xylene distillation) was between 4 and 5.5 percent for excelsior, and 6 and 7.5 percent for needles after preconditioning and before chemical treatment application.
Adjustments for Differences in Untreated Fuel Burning Rates
The burning characteristics of different batches of untreated fuels have varied somewhat in previous laboratory studies. Pine needles gathered in the fall have combustion rates that differ from those gathered in the spring, and fall needles may vary slightly from year to year. The same is true of batches of com- mercially prepared excelsior that may vary somewhat in un- treated flame spread rates and total weight-loss rate. Because of the variations in untreated fuels used in this study, adjustments were necessary to make burning data comparable. One method was to calculate the percentage that untreated fuel combustion rates for individual fires are reduced by each treatment. By this means, the differences in untreated burning rates (flame spread and fuel bed weight loss) are taken into account and numerical comparisons can be made. A 0 rating indicates that a retardant has no effect on combustion, and a 100 rating indicates that a chemical totally stops flaming and glowing combustion. The percentage rating is calculated using the flame-spread rate and weight-loss rate for untreated and treated fuels (pine needles and aspen excelsior). For comparisons to be valid, treatment levels for different fires or chemicals must be approximately equal. Spread- and weight-loss rates for treated fuels are calculated as percentages of spread- and weight-loss rates for untreated fuels (percent reduction). Equal numbers of fires for each treatment level (or weighted averages) must be used if statistically meaningful percent reduction factors are to be averaged for two or more treatment levels.
Another method was to adjust each treated fuel burning rate by a percentage corresponding to the differences in average un- treated burning rates for different batches of fuel. Recently used untreated fuels have burning rates varying from 5 to 17 percent higher than rates for untreated fuels used in the past when the DAP and M-MAP burns were conducted; therefore, in this study S-MAP burning rates have been adjusted down- ward. This adjustment also permits the use of the actual spread and weight-loss rates for computations and graphing, rather than conversion to percent reduction of untreated rates. Table 1 shows the untreated averages and the differences between needles and excelsior used in 1970 and 1980.
Table 1.—Factors for adjusting treated bed burning rates for differences in untreated burning rates
Aspen excelsior
Pine needles Average Std. Average Year RIS' dev. N RIW? Ft/min g/min 1980 2.03 0.10 4 345 1970 1.81 03: 12 294 Difference 0.22 51 0.22 51 1- 203 = 0.892 1 - 345 = 0.852 Adjustment 0.892 0.852
‘RIS = flame spread rate. 2R/W = burning fuel rate of weight loss.
Chemical Application
A pressurized sprayer with a fan-shaped spray pattern was used to apply retardant (George and Blakely 1972). The sprayer was calibrated for each different retardant chemical and con- centration to produce a spray pattern that would coat the fuels uniformly. The volume of chemical solution applied to each bed was held as closely as practical to 0.26 gal (1 liter), and the different levels of the dried chemical applied to different beds were varied by adjusting the solution concentrations. After treatment, fuel beds were dried under environmental conditions of 90° F + 2° (32.2° C + 2°), and 20 percent relative humid- ity + 2 percent, until all the solution water had been evaporated and the fuel moisture content was about the same as before treatment. Low-velocity fans were used to keep air moving above the treated beds so drying would occur uni- formly throughout the depth of the bed. After all the water had evaporated (determined by frequent weighing on an elec- tronic balance), beds were burned in a wind tunnel under con- ditions of 90° F (32.2°), 20 percent relative humidity, and ina 5-mi/h (8-km/h) wind. (These environmental conditions can be related to wildfire situations by the following: When needles and excelsior are classed as fuel type U, the National Fire- Danger Rating System [NFDRS] grades fires in untreated fuels as spread component 5, energy release rate 38, and burning in- dex 34.)
Std.
Average Std. Average Std. N RIS dev. N RIW dev. N Ft/min g/min 4.26 0.10 4 529 Son. 14 11 4.04 14° 9 459 6.1 9 0.22 70 0.22 70 1- 4.26 = 0.948 1- 529. = 0.868 0.948 0.868
Burning Procedures
A 3-ft- (0.91-m-) long untreated fuel bed of the same fuel type and loading as that in the treated bed was ignited and allowed to burn into the treated fuel. As the fuel burned, the rate of weight loss was continuously measured by four load cells mounted beneath the bed (George and Blakely 1970), and data were recorded on a Tektronix® 4051 microcomputer. The flame spread was monitored visually, and an event marker was used to record the flame front progress. After each fire, the recorded data were entered into a computer program, and flame spread rate and total fuel bed weight loss rate were calcu- lated and plotted. These two parameters were used for compar- ing the effectiveness of different chemicals and treatment levels.
RESULTS
About 250 treated and untreated beds were burned. Results of burning pine needles and aspen excelsior treated with DAP have been reported previously (George and Blakely 1972), and are used as a standard for USDA Fire Retardant Qualification Tests. The M-MAP-treated beds were burned during the same period as DAP-treated beds (George and Blakely 1972), and therefore untreated ponderosa pine and aspen excelsior from the same untreated fuel batches were used. (The M-MAP data have not been published previously.) The remaining fires were conducted using untreated fuels collected several years later. Adjustments were made (table 1) in the data for recent burns to compensate for differences in untreated burning rates. Tables for all MAP-treated (except M-MAP) beds contain col- umns of adjusted spread- and weight-loss rates that were used for regression analysis. These data and those for DAP and M-MATP are shown in tables 2 through 9.
Table 2.—Summary of test data for DAP-treated aspen excelsior fuel beds (from George and
Chemical by weight
Percent
Blakely 1972)
Treatment solutiom
Solution density
g/cm?
1.014 1.014 1.014 1.014 1.029 1.029 1.029 1.029 1.029 1.042 1.042 1.042 1.042 1.056 1.056 1.056 1.072 1.072 1.072
Solution quantity
g ml 1000 986 990 976 970 957 985 971 1030 1001 950 923 990 962 936 910 917 891 1000 960 980 940 970 931 1025 984 1000 947 995 942 1040 985 1030 961 1010 942 1005 938
Anhydrous chemical DAP P.O, ---- g/ft? ---- 2.08 1.12 2.06 1.11 2.02 1.09 2.05 1.10 4.29 2.31 3.96 2S 4.12 2.22 3.90 2.10 3.82 2.06 6.25 3.36 6.13 3.30 6.06 3.26 6.40 3.44 8.33 4.48 8.29 4.46 8.67 4.66 10.73 5.77 10.52 5.66 10.47 5.63
Rate of
flame
spread RIS
Ft/min
3.01 2.96 3.12 2.82 1.51
Rate of weight loss RIW
g/min
200 320 225 296 188 213 126
Table 3.—Summary of test data for DAP-treated ponderosa pine needle fuel beds (from George and Blakely 1972)
Treatment solutiom Anhydrous Rate of chemical flame Rate of Chemical Solution Solution a Spread weight loss
by weight density quantity DAP _—~PO, RIS RW Percent g/cm? g ml ---- gift? ---- Ft/min g/min 2.5 1.014 975 962 2.03 1.09 1.46 223 2.5 1.014 965 952 2.01 apa ks! 1.68 233 2.5 1.014 955 942 1.99 1.07 1:55 202 2.5 1.014 1000 986 2.08 1.12 1.70 196 5.0 1.029 1035 1006 4.31 2.32 1.56 205 5.0 1.029 995 967 4.15 2.23 1,27 160 5.0 1.029 990 962 4.12 2.22 1.08 169 5.0 1.029 1000 972 4.17 2.24 1.26 184 5.0 1.029 980 952 4.08 2.20 1.39 191 We5 1.042 1000 960 6.25 3.36 1.36 180 5 1.042 1010 969 6.31 3.39 1.21 164 te5 1.042 975 936 6.09 3.28 125 200 7.5 1.042 1025 984 6.41 3.45 .93 170 10.0 1.056 1020 966 8.50 4.57 82 163 10.0 1.056 995 942 8.29 4.46 69 174 10.0 1.056 1055 999 8.79 4.73 5 161 10.0 1.056 1010 956 8.42 4.53 94 196 12.5 1.072 1010 942 10.52 5.66 69 131 12.5 1.072 965 900 10.05 5.41 .70 147 2E5 1.072 1050 979 10.94 5.89 43 111 15.0 1.084 1025 946 12.81 6.89 .28 77 15.0 1.084 960 886 12.00 6.46 68 137 15.0 1.084 1000 923 12.50 6.73 50 129 15.0 1.084 990 913 12.38 6.66 soi) 101 17.5 1.108 1050 948 15.31 8.24 26 71 17.5 1.108 980 884 14.29 7.69 ROTA 94 17.5 1.108 1015 916 14.80 7.96 39 97 17.5 1.108 1025 925 14.94 8.04 38 79 20.0 ae Ea 1005 904 16.75 9.01 33 76 20.0 tells 1038 934 7-30 9.31 24 63 20.0 AAA 1081 973 18.02 9.69 .29 --
Table 4.—Summary of test data for M-MAP-treated aspen excelsior fuel beds
Treatment solutiom Anhydrous Rate of chemical flame Rate of Chemical Solution Solution a pices melgneaces by weight density quantity MAP P,0, RIS RW Percent g/cm? g ml ---- g/ft? ---- Ft/min g/min 2:5 1.014 1010 996 2.10 1.30 2.48 297 2.5 1.014 1010 996 2.10 1.30 2.63 294 2.5 1.014 995 981 2.07 1.28 2.23 264 5.0 1.029 1005 977 4.19 2.59 5 113 5.0 1.029 1135 1103 4.73 2.92 52 118 5.0 1.029 1055 1025 4.40 2.f2 1.06 133 7.5 1.042 1030 988 6.44 3.98 60 103 Ths5) 1.042 1100 1056 6.88 4.25 .68 79 7.5 1.042 1000 960 6.25 3.86 73 120 10.0 1.056 930 881 1.15 4.78 74 117 10.0 1.056 960 909 8.00 4.94 43 82 10.0 1.056 965 914 8.04 4.96 .40 83
Table 5.—Summary of test data for M-MAP-treated ponderosa pine needle fuel beds
Treatment solutiom
Chemical Solution by weight density
Percent g/cm? 25 1.014 25 1.014 2.5 1.014 5.0 1.029 5.0 1.029 5.0 1.029 7.5 1.042 15 1.042 7.5 1.042
10.0 1.056 10.0 1.056 10.0 1.056 12.5 1.072 12.5 1.072 12.5 1.072 12.5 1.072 12.5 1.072 12.5 1.072 15.0 1.084 15.0 1.084 17.5 1.108 Wes 1.108 17.5 1.108
Solution quantity
g ml 990 976 1035 1021 1040 1026 1035 1006 1090 1059 1045 1016 1070 1027 1060 1017 1070 1027 1100 1042 1135 1075 965 914 925 863 940 877 1040 970 1085 1012 1010 942 1115 1040 1125 1038 1060 978 1135 1024 1115 1006 1080 975
Anhydrous chemical MAP P.O; ----- gift? ----- 2.06 1.27 2.15 1.33 2.16 1.33
4.31 2.66 4.54 2.80 4.35 2.69 6.69 4.13 6.63 4.09 6.69 4.13 9.16 5.65 9.46 5.84 8.04 4.96 9.63 5.94 9.79 6.04 10.83 6.69 11.30 6.98 10.52 6.49 11.61 AWS 141 8.70 13.3 8.21 16.55 10.22 16.26 10.04 15.75 9.72
Rate of
flame
spread RIS
Ft/min
1.56 1.26
Table 6.—Summary of test data for S-MAP-treated aspen excelsior fuel beds
Treatment solutiom
Chemical Solution by weight density
Percent g/cm? 2.5 1.014 2.5 1.014 2-5 1.014 5.0 1.028 5.0 1.029 5.0 1.028 5.0 1.028 5.0 1.029 TS: 1.042 Wed 1.042 75 1.042
10.0 1.056 10.0 1.056 10.0 1.056
Solution
quantity g ml 992 978 988 974 1032 1018 1032 1004 1009 981 1041 1013 1076 1047 1022 994 1051 1009 1009 968 1034 992 1047 991 1061 1005 1046 991
Anhydrous chemical MAP P,0, ---- g/ft? ---- 2.06 1.28 2.06 1.27 2.15 1.33 4.30 2.65 4.20 2.60 4.34 2.68 4.48 2.77 4.26 2.63 6.57 4.05 6.31 3.90 6.46 3.99 8.73 5.39 8.84 5.46 8.72 5.38
‘Adjusted for difference in untreated fuel burning rate.
Rate of flame spread‘ RIS Ft/min 2.53 2.40 2.41 2.28 1.79 1.70
82 78 70 66
1.27 1.20 98 93 59 56 91 86 80 76 71 67 46 44 58 55 58 55
Rate of weight loss RIW
g/min
202 146 258 206 186
94 128 147
- Rate of weight loss’ RIW
g/min
216 187 223 194 171 148 136 118 247 214 134 116 223 194 124 108 137 136 126 109
117 102 58 50
Table 7.—Summary of test data for S-MAP-treated ponderosa pine needle fuel beds
Treatment solutiom
Chemical Solution by weight density
Percent g/cm* 2.5 1.014 25 1.014 2.5 1.014 25 1.014 2.9 1.014 5.0 1.028 5.0 1.028 5 1.042 thse) 1.042 1:5 1.042 7.5 1.042 75 1.042 7.5 1.042
10.0 1.056 10.0 1.056 10.0 1.056 A215 1.071 12.5 1.071 12.5 1.072 12.5 1.072 15.0 1.0384 15.0 1.084 15.0 1.084
‘Adjusted for differences in untreated fuel burning rates.
Solution quantity
g ml 1009 995 985 971 985 971 995 981 1006 992 1024 996 1044 1016 1024 983 1031 989 1016 975 1038 996 1060 1017 1024 983 1021 967 980 928 1051 995 1032 964 1032 964 1060 989 1062 991 1081 997 152 1063 1107 1021
Anhydrous chemical MAP P.O, ---- g/ft? ---- 2.10 1.30
2.05 1:27 2.05 2d 2.07 1.28 2.10 1.29 4.27 2.63 4.35 2.69 6.40 3.95 6.44 3.98 6.35 3.92 6.49 4.00 6.63 4.09 6.40 3.95 8.51 5125 8.17 5.04 8.76 5.41 10.75 6.64 10.75 6.64 11.04 6.82 11.06 6.82 13:51 8.34 14.40 8.89 13.84 8.54
Table 8.—Summary of test data for treated aspen excelsior fuel beds
Chemical Chemical treatment by weight
Percent
T-MAP 5.0
D-MAP 5.0
A-MAP 5.0
‘Adjusted for differences in untreated fuel burning rates.
Solution density
g/cm?
1.029 1.029 1.029 1.059 1.059 1.059
1.029 1.029 1.029 1.059 1.059 1.059
1.029 1.029 1.029 1.059 1.059 1.059
Treatment solutiom
Solution quantity g ml 1046 1017 1048 1018 1088 1057 1059 1000 1018 961 1010 954 1054 1024 1039 1010 1059 1029 1019 962 1024 967 1021 964 1028 999 1032 1003 1010 982 1045 987 1049 991 1021 965
Rate of flame Rate of spread' weight loss' RIS RIW Ftimin g/min 1.78 1.59 244 203 1.52 1.36 259 221 1.88 1.68 251 214 2.25 2.01 212 181 1.91 1.70 244 208 1.40 1.25 200 170 1.57 1.40 209 178 1.45 1.29 179 156 1.29 1.15 206 176 1.12 1.00 190 162 Val 99 138 118 1.18 1.05 129 110 1.07 95 162 138 93 83 167 142 94 84 129 110 86 hie 101 129 lie 64 159 135 82 73 = 136 116 59 53 83 (| A7 42 102 87 52 46 118 101 A8 43 108 92 1 sO 93 79 Anhydrous Rate of chemical flame Rate of spread’ weight loss' MAP P,O, RIS vy ---- g/ft? ---- Ft/min g/min 4.36 2.69 1.01 .96 236 205 4.37 2.70 94 89 192 167 4.53 2.80 83 79 187 162 8.83 5.45 39 37 97 84 8.48 5.24 59 56 110 95 8.42 5.20 52 49 108 94 4.39 2.01 89 84 -- -- 4.33 2.67 91 .86 178 155 4.41 212 .60 57 198 172 8.49 5.24 40 38 92 80 8.53 5:27 40 38 84 73 8.51 5.25 54 51 95 82 4.28 2.64 .98 93 156 135 4.30 2.65 ahi 68 144 125 4.21 2.60 AQ 46 92 80 8.71 5.38 33 31 88 76 8.74 5.40 43 41 134 116 8.51 525 54 51 85 74
Table 9.—Summary of test data for treated ponderosa pine needle fuel beds
Treatment solutiom
Chemical Chemical Solution Solution treatment by weight density quantity Percent g/cm? g ml
T-MAP 5.0 1.029 1041 1012 5.0 1.029 1033 1004
5.0 1.029 1034 1005
10.0 1.059 1005 949
10.0 1.059 1032 975
10.0 1.059 1026 967
D-MAP 5.0 1.029 1008 980 5.0 1.029 1050 1021
5.0 1.029 1028 1000
10.0 1.059 1033 976
10.0 1.059 1053 995
10.0 1.059 1024 967
A-MAP 5.0 1.029 1037 1008 5.0 1.029 1003 975
5.0 1.029 1040 1011
10.0 1.059 1044 986
10.0 1.059 1024 967
10.0 1.059 1026 969
‘Adjusted for differences in untreated fuel burning rates.
In the George and Blakely (1972) study, DAP was tested at several treatment levels, and regression equations were deter- mined for flame spread and weight loss rates on pine needles and aspen excelsior fuels. The same type regression analysis was used with M-MAP and S-MAP, and regressions for all three chemicals have been compared. The analysis was performed to determine if differences exist between the fire-retarding effec- tiveness of the three source-samples of P,O;. To perform statis- tical tests, it was assumed that there was no significant differ- ence in overall effectiveness when equal levels of P.O; were ap- plied. The hypothesis was tested by covariance analysis and an saettests
Rates of flame spread and fuel weight loss (energy release) were fitted by a least-squares method to determine what equa- tion form (quadratic, exponential, logarithmic, reciprocal, and so forth) would fit best and give high correlation coefficients.
Anhydrous Rate of chemical flame Rate of spread’ weight loss’ MAP PO, RIS RIW ---- g/ft? ---- Ft/min g/min 4.34 2.68 1.54 We87/ 188 160 4.40 2.66 1.25 Ve 172 147 4.31 2.66 1.33 1.19 204 174 8.38 5.17 82 73 140 119 8.60 5.31 1.09 97 158 135 8.55 5.28 .98 87 126 107 4.20 2.59 1.65 1.47 190 162 4.38 2.70 1.46 1.30 199 170 4.28 2.64 1:51 1.35 233 190 8.61 5.31 Uciks} 1.01 -- -- 8.78 5.42 1.07 95 160 136 8.53 5.27 84 a5 190 162 4.32 2.67 1.67 1.49 169 144 4.18 2.58 Ueree/ 1.58 190 162 4.33 2.67 1.44 1.26 202 172 8.70 5.37 1.01 .90 181 154 8.53 5.27 1.00 89 168 143 8.55 5.28 .93 83 203° 173
Some of the best-fit equations shown in tables 10 and 11 do
not have the highest R? values possible because data groups that were tested against each other required that their regression equations be of the same form (for example, all ponderosa pine rate-of-spread data are in a second-degree polynomial form so that ‘‘F’”-tests can be performed). (Equation use is limited to the range included in the data sets and extrapolations beyond the real data cannot be expected to predict accurately.) Three individual and three paired equations were formed with data for each chemical for each test parameter. Then data for all three chemicals (triplet) for each parameter were pooled, and another best-fit equation was formed. Each paired equation was tested against the triplet equations and each individual against each other individual equation by an ‘‘F’’ test method described in figure 1.
Table 10.—Regression equations for flame-spread rate and weight-loss rate for ponderosa pine needles
Treatment N Equation R? F! Significance?’ Variance ratio Percent
Rate of spread
DAP 31 Y = 1.9420 — 0.30426X + 0.01313X? 0.93
M-MAP 23. Y = 1.9419 — 0.31906X + 0.01543xX? 81
S-MAP 24 Y = 2.0152 — 0.29478X + 0.01255xX? 91 Pooled
DAP/M-MAP 54 Y = 1.9523 — 0.31616 + 0.01471X? 88 0.138 (3, 48) NS VS All-pooled 76 (3, 48) NS
DAP/S-MAP 55 Y = 1.97073 — 0.29478X + 0.01213X? 91 2.36 (3, 49) NS VS/All-pooled 2.64 (3, 49) NS
M-MAP/S-MAP 47 Y = 1.99750 — 0.31072X + 0.01410X? 86 1:33 (3, 41) NS VS All-pooled 1.55 (3, 41) NS
DAP/M-MAP/S-MAP 78 Y = 1.97724 — 0.30872X + 0.01377X? 88
Rate of weight loss
DAP 30 Y = 234.18 — 18.002X 87
M-MAP 22 Y = 214.02 — 15.693X 63
S-MAP 24 Y = 217.77 — 16.471X .80 Pooled
DAP/M-MAP 52 Y = 227.53 — 17.280X ATA 97 (2, 48) NS VS All-pooled puts (2, 48) NS
DAP/S-MAP 54 Y = 226.56 — 17.261X 83 1.94 (2, 50) NS VS All-pooled 2.09 (2, 50) NS
M-MAP/S-MAP 46 Y = 215.99 — 16.053X ihe 11 (2, 42) NS VS All-pooled ‘57 (2, 42) NS
DAP/M-MAP/S-MAP 76 Y = 223.83 — 16.915X 718
‘Test for the reduction in variance between pooled and unpooled models. 2All differences in regressions are not significant below the 99 percent level.
Table 11.—Regression equations for flame-spread rate and weight-loss rate for aspen excelsior
Treatment
Rate of spread DAP
M-MAP S-MAP
Pooled
DAP/M-MAP
VS All-pooled DAP/S-MAP
VS/All-pooled M-MAP/S-MAP
VS All-pooled DAP/M-MAP/S-MAP
Rate of weight loss
DAP M-MAP S-MAP
Pooled
DAP/M-MAP
VS All-pooled DAP/S-MAP
VS All-pooled M-MAP/S-MAP
VS All-pooled DAP/M-MAP/S-MAP
N
19
12
14
31
33
26
45
19
12
12
31
31
24
43
<<<
<<<
Y
v
I
Equation R? F' Significance?
Variance ratio Percent
— 0.33231 + 3.5877 (X~') 0.97
— 0.27498 + 3.4232 (X—') 92
— 0.05014 + 2.7322 (X~') 87
— 0.31400 + 3.5369 (X~') 95 0.14 (2, 27) NS 1.12 (2, 27)
— 0.24474 + 3.3264 (X—') 93 4.19 (2, 29) 95 4.21 (2, 29) 95
— 0.16302 + 3.0786 (X~') 89 1.40 (2, 22) NS 2.23 (2, 22) NS
— 0.25226 + 3.3512 (X~ ') 93
264.86 — 121.24 (Ln X) 81
306.48 — 147.31 (Ln X) 89
196.25 — 52.689 (Ln X) 37
278.43 — 129.26 (Ln X) 83 1.36 (2, 27) NS 3.20 (2, 27) NS
241.43 — 96.375 (Ln X) 65 3.74 (2, 27) 95 4.38 (2, 27) 95
249.46 — 97.87 (Ln X) 61 6.13 (2, 20) 99 6.57 (2, 20) 99
257.21 — 108.64 (Ln X) 70
'Test for the reduction in variance between pooled and unpooled models. ?All differences in regressions are not significant below the 99 percent level.
Chemical 1, N, points f, (x) = a, + a,X + a,X? . . a,x?
Chemical 2, N, points f, (x) = b, + b,X + b,X? . . bLXP
Combine 1 and 2, N, + N, points f, (x) = ¢, + ¢,X + CX? . . ¢,XP
Note: All three regressions must be of the same form; i.e., log, third degree polynomial, and so forth.
A N, SSYs4— y (y,i — f,()) ie—wal A N, SSY, = r (y,i — f,(x;)?) ital Zl (N, + N,) SS) (= y (yi — f,(x;))?
i=
SS¥= SsY, - Sssy, = difference in SSY
(N, + N, — P) — (N, — P) -(N, — P) =0p
(SSY, + SSY,)/(N, + N, — P) = MSE(Y) Diff SSY pas alt MSE(Y)
F (p, (N,+ N,— 2)) =
Figure 1.—Method used to calculate F values.
10
The covariance analysis shows no significant difference be- tween the fire-retarding effectiveness of any of the three chemicals except for one parameter. There appears to be a statistical difference between the regression for M-MAP and S-MAP on excelsior weight loss. The greatest differences are between data at the low treatment levels where small variations in treatment amounts or fuel moisture content will sometimes cause large-scale differences. (The curves show that fire retard- ant effectiveness is very sensitive to small changes within the low-treatment areas.) Examination of the M-MAP and S-MAP regression data shows that their curves cross at about the 3-g/ft? (929-cm?) treatment level. Apparent differences within each separate regression for each chemical can be caused by variation in fuel chemical composition, environmental condi- tions, and retardant application. The apparent differences be- tween the fire-retarding abilities of DAP, M-MAP, and S-MAP are because of these variations in testing procedures; and the apparent differences in total effectiveness are therefore not real but because of experimental error. When the three chemicals, two test parameters, and two fuel types are all combined, analysis indicates no real significant differences exist (at the 0.05 level) between the fire- retarding abilities of DAP, M-MAP, and S-MAP.
Two methods were used to evaluate D-MAP, A-MAP, and T-MAP. One method was to plot the individual points on the curves for pooled DAP/M-MAP/S-MAP (P.O; curves) and make general comparisons. The other method was to calculate the flammability reduction percentage for each chemical and compare it to the reduction calculated for the same level of P.O, from the pooled DAP/M-MAP/S-MAP equation. Tables 12 through 14 show the percent reductions for the different chemicals compared to reductions for pooled P,O;. The reduc- tions compare closely for all chemicals with none varying more than 0.01 from the pooled P.O; regression.
Table 12.—Reduction of untreated’ fuel combustion rates caused by D-MAP treatment!
R/S R/IW Cumulative
% reduction RIW % reduction % reduction P,0, R/S 1 2 1 2 1 2 g/ft? Ft/min g/min
Ponderosa pine
2.59 1.47 0.19 0.30 162 0.45 0.39 2.70 1.30 .28 31 170 42 39 2.64 lpsttoy seds) 30 190 230) 39 5:31 1.01 44 60 5.42 95 48 61 136 54 55 5.27 TAS) a ene) .60 162 45 54 Average 0.37 0.45 0.44 0.45 0.41 0.45 Aspen excelsior 2.71 0.84 0.79 0.76 2.67 86 .79 he) 155 0.66 0.67 2.72 57 ~~ .86 16 172 63 68 5.24 38 © =6.91 .90 80 83 83 5.27 38) | 9/1 .90 78 85 83 5.25 51 87 .90 82 83 83 Average 0.86 0.83 0.74 0.75 80 19
Total average 0.61 0.62
‘4 = percent reduction of combustion parameters (flame spread rate and weight loss rate) for untreated fuel bed rates.
2 = percent reducton of combustion parameters that are computed using equivalent treatment amounts and data (flame spread rate or weight loss rate) from pooled data curves for DAP, M-MAP, and S-MAP.
Table 13.—Reduction of untreated fuel combustion rates caused by A-MAP treatment’
RIS RIW Cumulative % reduction RIW % reduction % reduction P,0, RIS 1 2 1 2 1 2 g/ft? Ft/min g/min Ponderosa pine 2.67 1.49 0.18 0.31 144 0.51 0.39 2:58) OBE saves 30 162 45 .39 2.67 1.26 .30 oi 172 42 39 5:37, 901 + .50 60 154 A8 55 527, 89 51 .60 143 51 54 5.28 283,62 '54 .60 173 1 54 Average 0.36 0.45 0.46 0.47 0.41 0.46 Aspen excelsior 2.64 0.93 0.77 0.75 135 0.71 0.67 2.65 68 83 74 125 0.73 0.67 2.60 46 89 74 80 83 67 5.38 roll 92 91 76 84 84 5.40 41 .90 91 116 75 84 5.25 51 .87 90 74 84 83 Average 0.86 0.83 0.78 0.75 82 79 Total average 0.62 0.63
‘1 = percent reduction of combustion parameters (flame spread rate and weight loss rate) for untreated fuel bed rates.
2 = percent reducton of combustion parameters that are computed using equivalent treatment amounts and data (flame spread rate or weight loss rate) from pooled data curves for DAP, M-MAP, and S-MAP.
11
Table 14.— Reduction of untreated fuel combustion rates caused by T-MAP treatment’
R/S RIW Cumulative
% reduction R/W % reduction % reduction
P,0, R/S 1 2 1 2 1 2 g/ft? Ft/min g/min
Ponderosa pine
2.68 1.37 0.24 0.31 160 0.46 0.39 2.66 ten este! 31 147 50 39 2.66 1.19 34 a3il 174 AA 39 5.17 73) “60 59 119 60 54 5.31 97 ~=.46 .60 135 54 54 5.28 87) 202 .60 107 64 54 Average 0.42 0.45 0.53 0.47 0.48 0.46 Aspen excelsior 2.69 0.96 0.76 0.75 205 0.55 0.67 2.70 89 .78 15 167 0.64 0.68 2.80 19 ~~ 80 “Ml, 162 65 68 5.45 le Oil 91 84 82 84 5.24 256n 3O .90 95 19 83 5.20 49 88 90 94 80 83 Average 0.83 0.83 0.71 0.75 AT 19 Total average 0.63 0.63
‘4 = percent reduction of combustion parameters (flame spread rate and weight loss rate) for untreated
fuel bed rates. 2 = percent reducton of combustion parameters that are computed using equivalent treatment amounts and data (flame spread rate or weight loss rate) from pooled data curves for DAP, M-MAP, and S-MAP.
Tables 10 and 11 show the results of regression analysis and ‘*F’’ tests for regression differences. The program for calculating ‘‘F’’ values is given in figure 1. Figure 2 shows all the weight-loss data on excelsior for DAP, M-MAP, and S-MAP, and the pooled equation is plotted. Figure 3 shows data and the equation for spread rate data on excelsior for DAP, M-MAP, and S-MAP. Figures 4 and 5 are weight-loss and spread-rate data on pine needles for all three chemicals. Figures 6 through 9 show all data points for all three chemicals, and individual best-fit equations are plotted for each chemical.
12
+ — DAP 456 o — M-MAP
% — S-MAP
RATE OF WEIGHT LOSS Cg/m)>
P,O, g/sq. ft.
Figure 2.—Effect of DAP, M-MAP, and S-MAP on weight loss rate of excelsior. (Equation for pooled data is plotted.)
RATE OF SPREAD Cft/m)>
P,@, Cg/ft sq)
Figure 3.—Effect of DAP, M-MAP, and S-MAP on spread rate of excelsior. (Equation for pooled data is plotted.)
13
358
388
258
288
{Sa
188
RATE OF WEIGHT LOSS Cg/m)>
14)
P,O; Cg/ft sqd
Figure 4.—Effect of DAP, M-MAP, and S-MAP on weight loss rate of pine needles. (Equation for pooled data is plotted.)
RATE OF SPREAD Cft/m>
4) 1 Z S 4 S 6257. 8 9 18 11 P,O; Cg/Ft sqd
Figure 5.—Effect of DAP, M-MAP, and S-MAP on spread rate of pine needles. (Equation for pooled data is plotted.)
14
RATE OF SPREAD Cft/m)>
P,Os Cg/Ft sq>
Figure 6.—Effect of DAP, M-MAP, and S-MAP on spread rate of excelsior.
+ — DAP 458 o — M-MAP % — S-MAP
480
SYS) € NG (0)
Y¥ 3288 ” O
re 250 | 5
H 288 Ww =
BS 150 WW |
~ 108
4)
Q
Pj 0p Cg7eq tf tD
Figure 7.—Effect of DAP, M-MAP, and S-MAP on weight loss rate of excelsior.
15
358 l + — DAP
o — M-MAP * — S-MAP
250) ie
288
{58
{88
RATE OF WEIGHT LOSS Cg/m)>
P,O, Cg/Ft sq>
Figure 8.—Effect of DAP, M-MAP, and S-MAP on weight loss rate of pine needles.
RATE OF SPREAD Cft/m>
P,0- Cg/Ft sad
Figure 9.—Effect of DAP, M-MAP, and S-MAP on spread rate of pine needles.
16
SUMMARY AND DISCUSSION
The study objective was to compare the fire-retarding effec- tiveness of diammonium phosphate and several samples of monoammonium phosphate. Diammonium phosphate has been the standard for several years; therefore, the other chemicals were compared to its effectiveness. M-MAP and S-MAP proved to be as effective as DAP for retarding flaming and glowing combustion when compared on an equal P20; equiv- alent basis under statistical analysis. The other three samples—D-MAP, A-MAP, and T-MAP—(even though fewer burning tests were performed) appear to be equally as effective. The differences in flammability reduction are because of experi- mental error and are not statistically significant. They are prob- ably caused by inconsistencies and variations in fuel bed con- struction, fuel physical configuration, fuel moisture content, environmental conditions, and so forth.
These tests and others (George and Susott 1971) indicate that the most important chemical characteristic is the available phosphorus (P). As the chemical is heated, the phosphate (PO,) compounds are converted to phosphoric acid (H;PO.,) that alters pyrolysis of the fuel. Both diammonium and monoam- monium phosphates are converted easily to H,PO,because the ammonia cations (NH;) are driven off at low temperature:
166° C (NH.)2 HPO, > NH, H,PO, + NH; ¢ 190°C A H;PO, + NH;t >H,0 + P20;
Phosphate anions (PO,), when combined with sodium (Na), calcium (Ca), potassium (K), and others, cannot be converted to H;PO, readily, and therefore do not make the PO, available in the most effective form as fire retardant.
Whether or not one or two ammonias are associated with the P does not appear to make a difference. The method for asso- ciating the ammonia with the phosphate (PO,) also does not appear to affect the fire-retardant ability. Whether ammonia is extracted from coal smoke that is being ‘‘scrubbed’’ with phosphoric acid (D-MAP) or whether the acid is being am- moniated by bubbling ammonia gas into it (S- and T-MAP) seems to make no difference in the availability of P and the
resulting fire-retarding effectiveness. A-MAP, produced from a less pure acid, is as effective as the other MAP forms when most of the impurities have been removed after ammoniation. The fire-retarding effectiveness of each MAP (and also DAP), when in a pure form, can be equated on the P or P,0; content. Any formulations containing impurities may change the level of effectiveness.
PUBLICATIONS CITED
Brown, F. L.; Tang, W. K. Effect of various chemicals on thermogravimetric analysis of ponderosa pine. Res. Pap. 6. Madison, WI: U.S. Department of Agriculture, Forest Serv- ice, Forest Products Laboratory; 1963. 20 p.
Eickner, H. W. Basic research on the pyrolysis and combustion of wood. For. Prod. J. 12(3): 194-199; 1962.
George, C. W. Liquids fight forest fires. Fert. Solutions. 15(6): 10-11, 15, 18, 21; 1971.
George, C. W.; Blakely, A. D. Energy release rates in fire retardant evaluation. Fire Technol. 6(3): 203-210. 1970.
George, C. W.; Blakely, A. D. Effects of ammonium sulfate and ammonium phosphate on flammability. Res. Note INT-121. Ogden, UT: U.S. Department of Agriculture, For- est Service, Intermountain Forest and Range Experiment Sta- tion; 1972.-26 p.
George, C. W.; Susott, R. A. Effects of ammonium phosphate and sulfate on the pyrolysis and combustion of cellulose. Res. Note INT-90. Ogden, UT: U.S. Department of Agricul- ture, Forest Service, Intermountain Forest and Range Experi- ment Station; 1971. 27 p.
George, C. W.; Blakely, A. D.; Johnson, G. M.; Simmerman, D. G.; Johnson, C. W. Evaluation of liquid ammonium polyphosphate fire retardants. Gen. Tech. Rep. INT-41. Ogden, UT: U.S. Department of Agriculture, Forest Service, Intermountain Forest and Range Experiment Station; 1977. 54 p.
Hardy, C. E.; Rothermel, R. C.; Davis, J. B. Evaluation of forest fire retardants—a test of chemicals on laboratory fires. Res. Pap. INT-64. Ogden, UT: U.S. Department of Agricul- ture, Forest Service, Intermountain Forest and Range Experi- ment Station; 1962. 33 p.
Johansen, R. W. Monoammonium phosphate shows promise in fire retardant trials. Res. Note SE-137. Macon, GA: U.S. Department of Agriculture, Forest Service, Southeastern Forest Experiment Station; 1959. 2 p.
Johansen, R. W.; Shimmel, J. W. Increasing the viscosity of water and chemical fire retardants with clays and gums. Res. Pap. 19. Macon, GA: U.S. Department of Agriculture,
Forest Service, Southeastern Forest Experiment Station; 1963.7 p.
Johansen, R. W.; Crow, G. L. Liquid phosphate fire retardant concentrates. Fire Contr. Notes. 26(2): 13-16; 1965.
Miller, H. R.; Wilson, C. C. A chemical fire retardant—results of field trials using sodium calcium borate on forest fires in 1956. Tech. Pap. PSW-15. Berkeley, CA: U.S. Department of Agriculture, Forest Service, Pacific Southwest Forest and Range Experiment Station; 1957. 11 p.
Operation Firestop. Progress report No. 4, retardants - I. Berkeley, CA: U.S. Department of Agriculture, Forest Serv- ice, Pacific Southwest Forest and Range Experiment Station; 195Sa. 12 p.
Operation Firestop. Progress report No. 9, aerial firefighting. Berkeley, CA: U.S. Department of Agriculture, Forest Serv- ice, Pacific Southwest Forest and Range Experiment Station; 195Sb. 13 p.
Phillips, C. B.; Miller, H. R. Swelling bentonite clay—a new forest fire retardant. Tech. Pap. PSW-37. Berkeley, CA:
U.S. Department of Agriculture, Forest Service, Pacific Southwest Forest and Range Experiment Station; 1959. 30 p.
Truax, T. R. The use of chemicals in forest fire control. Rep. 1199. Madison, WI: U.S. Department of Agriculture, Forest Service, Forest Products Laboratory; 1939. 11 p.
Tyner, H. D. Fire extinguishing effectiveness of chemicals in water solution. Ind. Eng. Chem. 33(1): 60-66; 1941.
Wood, W. C. Liquid fertilizer tested as fire retardant. Fire Contr. Notes 31(2): 3-5; 1970.
17
Blakely, A. D. Monoammonium phospnate: effect on flammability of excelsior and pine needles. Res. Pap. INT-313. Ogden, UT: U.S. Department of Agriculture, Forest Service, Intermountain Forest and Range Experiment Station; 1983. 17 p.
The study quantified differences between fire-retarding abilities of monoammonium phosphate samples from five different sources. Ponderosa pine needles and aspen excelsior fuel beds were spray-treated with different levels of chemical solutions, dried, and burned under controlled laboratory conditions. Flame spread and energy release rates were used for comparisons. All five monoammonium phosphate samples proved to be equally effective.
KEYWORDS: forest fire retardent, ammonium phosphate, monoammonium phosphate, flame spread rate, energy release rate, chemical manufacturing method
The Intermountain Station, headquartered in Ogden, Utah, is one of eight regional experiment stations charged with providing scientific knowledge to help resource managers meet human needs and protect forest and range ecosystems.
The Intermountain Station includes the States of Montana, Idaho, Utah, Nevada, and western Wyoming. About 231 million acres, or 85 percent, of the land area in the Station territory are classified as forest and rangeland. These lands include grasslands, deserts, shrublands, alpine areas, and well-stocked forests. They supply fiber for forest industries; minerals for energy and industrial development; and water for domestic and industrial consumption. They also provide recreation opportunities for millions of visitors each year.
Field programs and research work units of the Station are main- tained in:
Boise, Idaho
Bozeman, Montana (in cooperation with Montana State Univer- sity)
Logan, Utah (in cooperation with Utah State University)
Missoula, Montana (in cooperation with the University of Montana)
Moscow, Idaho (in cooperation with the University of Idaho) Provo, Utah (in cooperation with Brigham Young University)
Reno, Nevada (in cooperation with the University of Nevada)
yz U.S. GOVERNMENT PRINTING OFFICE: 1983—676-032/1009 REGION No.8