Fire fighters are usually the first to arrive at the scene of a chemical incident. With the foams they carry, they can cover a liquid source of toxic vapors in short time, and thus, mitigate the toxic hazards drastically [US Army SBCCOM, 2003]. In this article we describe a method to standardize measurements that assess the use of fire fighter foams and the effectiveness of decontamination foams against chemical warfare agents (CWA). The same method can be used on CWA simulants and can be used in training. Our results indicate that an initial foam cover of 10 cm thickness can contain vapors for several hours. In general, the application of foam does not spread out the chemicals, spraying of any foam is better than no foam, as long as there a violent reaction is not expected. In tests with GB, the best results have been obtained with the decontamination foam CASCAD [Allen-Vanguard, 2009; Aue, Guidetti, 2002] using a 10 cm layer of foam that contains enough active chlorine to destroy GB.
In a case of chemical terrorism, first responders might well be confronted with a liquid source of toxic vapor that continues to spread its hazardous contents, including with vaporization. With foam, which is always available to fire fighters, and which is an efficient and simple means, such a source could be covered in seconds and the spread of vapors mitigated drastically. Once covered, the source could then wait for a longer time to be removed carefully and professionally by a decontamination team.
In order to evaluate foams that could be useful for covering toxic vapor sources, a large set of measurements has been performed in order to answer the following questions:
- Which foams could be used for this purpose?
- How thick should the foam cover be?
- For how long would such a foam cover be effective?
- Could the practical application of foam cause a spread of the toxic chemical?
The basic idea behind the experiments was to measure the time to breakthrough of CWA agents and simulants as a function of layer thickness for a selection of fire fighting and other foams. Additional tests were run in order to assess the possible spread of toxic chemicals by the application of foam.
MATERIALS AND METHODS
In order to obtain consistent results for breakthrough times, the following laboratory model was developed:
Foam thickness measurements were done in cylindrical glass beakers of 60 mm inner diameter (i.d.) and heights of 10, 20, 30, 50, 70, and 100 mm. A sample of 0.5 mL of each chemical agent was pipetted into small dishes of 37 mm i.d. and 6mm height on the bottom of the beakers in order to avoid special effects between foam and beaker walls affecting the agent.
In order to simulate sprayed foam, foam was sprayed from a distance of 2 m into a 60 cm x 60 cm x 20 cm enamel-coated metal tub mounted with one corner pointing down and tilted, such that the foam would flow out of the tub. The foam was transferred into the measuring beakers to cover the agent, using either a lab beaker or “pastry glazing” bag with a nozzle of 14 mm diameter, depending on how readily the foam flowed (for thinner foams, the lab beaker was used; the “pastry glazing” bag was used for thicker foams). After filling the beakers, excess foam was removed by leveling with a straight-edged ruler to obtain the proper foam thickness.
The chemical vapor evaporating through the foam was determined using a “chemical measurement lid”, described as follows: the measuring beaker was covered with a glass lid, which had an inner height of 12 mm and four vents spaced equally on its circumference. Three equal air streams of 80 L/h in total were fed through three of the vents, in order to clear the vapor space evenly through the fourth vent. A T-tube was attached to the fourth vent. This arrangement of tubes and vents is shown in [Figure 1]. One end of the T-tube was open, and, a chemical monitoring device was attached to the other end. The airflow through the vapor space was larger than the maximum intake flow of any of the monitors. In this way, the flow through the vapor space, and accordingly the dilution of the diffusing agent vapors, was kept constant irrespective of the intake of the monitor.
The breakthrough threshold concentration was arbitrarily set at 200 µg/m3 for HD and 100 µg/m3 for the other agents. Finding a chemical agent monitor sensitive enough to the agents, but not cross sensitive to the foams was quite a challenge. In the end, the monitors used, were Proengin AP2C, Smiths Detection Swiss CAM, Bruker RAID M-100, and for the Jomos protein foam the Inficon Hapsite Smart II. On the latter, the cycle time could be minimized to 10 min. Each detector was calibrated appropriately over the respective foams.
The effect of different foam thicknesses was measured at the same times after covering CWA. The same volume of CWA, 0.5 mL, was placed in each of successively taller beakers. The CWA in each beaker was covered at the same time with foam from the same batch. To measure the thinnest foam layer, the “chemical measurement lid” was put onto the shortest beaker, until breakthrough was measured. Then, it was placed on the next thicker foam layer in the next higher beaker and the breakthrough time was measured. This method works assuming breakthrough times are longer for thicker foam layers.
In order to characterize foam quality, the separation (dehydration) of each foam was measured as a function of time, in a 1 L graded conical glass sedimentation vessel. This is shown in [Figure 2]. The vessel was not covered, and, the influence of the ambient humidity changes was deemed unimportant. The characteristic parameter, namely the time needed to reach 90% of the final liquid volume, i.e., D90, was calculated by means of a fitting program. These dehydration measurements were done in parallel with the breakthrough measurements, which proved to be a crucial in getting credible and consistent results. We assumed that the dehydration times did not change when the foam was in contact with a CWA. All measurements were done twice at an ambient temperature of 20°C.
In real life situations, there has been concern that toxic chemicals could be spread out by the application of foam, and consequently a question as to whether it would not be better to leave the toxic source uncovered. In order to tackle this problem, an additional experiment was performed: In an 80 m3 closed climate chamber with recirculating air at a speed of 1-2 m/s, five different foams were applied to a layer of 0.2 mm of simulant CWA (MS or methyl salicylate) in a tub and from a distance of 3 m [Figure 3].
For CASCAD and MXOL foams (see below), a 60 cm x 60 cm x 20 cm tub was used; for the other foams, a smaller (25 cm x 35 cm x 11 cm) tub was used because of the small size of the foam generator. The concentration of simulant in the air was monitored with a Swiss CAM and the amount of the chemical evaporated integrated over time was reported. The different sizes of the tubs, and consequently different strengths of the sources, were accounted for in the evaluation.
The foams were selected from a cross section of fire extinguisher foams available in Switzerland, and included the decontamination foam CASCAD, introduced in the Swiss army for the NBC troops. In detail, they were:
- CASCAD: aqueous decontamination foam with active chlorine at pH = 10,
Allen-Vanguard Corp., 2400 St Laurent Blvd, Ottawa, ON K1G 6C4; Canada;
- 3M FC 602 ATC plus: AFFF lightwater foam at pH = 6,
3M Corporate Headquarters, 3M Center, St. Paul, MN 55144-1000, USA;
- Jomos Schaumgeist: protein foam at pH = 6,
JOMOS Brandschutz AG, Sagmattstrasse 5, CH-4710 Balsthal, Switzerland;
- Minimax MXOL 09 at pH = 4.5,
Minimax AG Bern, Dorfstrasse 10, CH-3075 Rüfenacht, Switzerland;
- Moussol: alcohol resistant AFFF,
Dr. R. Sthamer GmbH, Liebigstrasse 5, D-22113 Hamburg, Germany;
- Primus LS TS: ATC air foam at pH = 8,
Primus AG, Bottmingerstrasse 70, CH-4102 Binningen, Switzerland;
- Primus LW TN: ATC Light water foam,
Primus AG, Bottmingerstrasse 70, CH-4102 Binningen, Switzerland;
- Solberg Arctic foam 600 ATC at pH = 5.5,
Solberg Scandinavian AS, Olsvollstranda, N-5938 Saebovagen, Norway;
- Solberg Rehealing foam RF 3x6 at pH = 6.5,
Solberg Scandinavian AS, Olsvollstranda, N-5938 Saebovagen, Norway.
In bold are the names of the foams as they will be used for the remainder of the paper.
CASCAD was prepared in the dedicated backpack set according to the company's procedures with a pressure of 6 bar. MXOL and Primus foams were applied from regular fire extinguishers. FC 602 (6%), Jomos (3%), Moussol (3%), Arctic (6%) and Rehealing (6%) foams were prepared by respective dilutions of concentrates with deionized water and applied with a small foam generator at a pressure of 6 bar, corresponding to the foam pressure found on fire fighter trucks.
Chemical warfare agents and simulants
For the glass beaker experiments, breakthrough times were measured for several simulants. At an early stage of the project, when breakthrough was assessed by nose, banana oil and methyl salicylate (MS) were the simulants of choice; later on, when using the chemical monitors, DMMP was also used. As for chemical warfare agents, GB was evaluated with four foams, whereas GD and HD were evaluated with CASCAD, only. The purity of the simulants was technical grade; GB, GD and HD were better than 95% pure. For the measurements in the climate chamber, MS was used, exclusively.
The essential results are assembled according to the four main topics:
- breakthrough times; how long the foam retards the vaporization of the agent
- correlation of breakthrough times compared with dehydration times
- correlation of breakthrough times of DMMP simulant with GB
- spreading of chemical by application of foam
The time course of a breakthrough experiment is illustrated in [Figure 4]. The threshold concentration for breakthrough was chosen to be 100 µg/m3 of GB, corresponding to 22 µg/m3 of phosphorus measured with the detector. The concentration of phosphorus in the vapor space above the foam, i.e., the breakthrough of GB, is shown as a function of time using CASCAD foam, with foam layers of 3, 5, 7 and 10 cm. The blue peak is the breakthrough when the foam layer is 3 cm and the red peaks for 5 and 7 cm, respectively. When 10 cm of foam was used, there was no breakthrough.
The breakthrough times as a function of foam layer thickness can be determined from diagrams such as shown in [Figure 4]. These breakthrough times are the key results of these experiments and are shown in [Figure 5] for the four longest lasting foams: CASCAD (DMMP, GB, GD and HD), FC602 (GB and DMMP), Reheal (GB and DMMP) and JOMOS (GB only).
Five features are obvious from [Figure 5]:
- The breakthrough times rise with increasing foam layer thickness.
- The breakthrough times fall in the foam sequence CASCAD - Jomos - FC 602 - Rehealing.
- For the same foam, the breakthrough times are clearly shorter for the highly volatile GB than for DMMP.
- According to the nature of experiments with foam, there is considerable scatter in the data.
- A 10 cm layer of CASCAD foam contains enough active chlorine to destroy 0.5 mL of GB.
In essence, the barrier effect of foam lasts as long as the foam contains some quantity of water. As an illustration of this fact, the relation of dehydration times D90 and breakthrough times for 5 cm layers of CASCAD, Jomos, Rehealing and FC 602 is illustrated for GB in [Figure 6].
The clean correlation in [Figure 6] indicates that assessing D90 from the dehydration process is an excellent means to monitor the quality of the different foams. If the quality of the foam is good, meaning that it does not separate or dehydrate quickly or inconsistently with respect to other foams, then its behavior with any particular agent will be consistent. This measure lets us detect problems in foam production at an early stage and without the expense of using chemical agents. Furthermore, we can conclude that the easily measured dehydration times D90 yield a good forecast for breakthrough times.
The breakthrough times of DMMP and GB have been correlated in pairs for the same foam and the same layer thickness for a series of layer thicknesses and three foams. This should permit the investigation of whether the vapors of the two agents are affected differently by the three foams and, therefore the influence of the different chemistries of the three foams can be noted. The correlation of breakthrough times through different layer thicknesses of a selection of 3 foams is shown in [Figure 7].
The data in [Figure 7] show that the different foams confined the vapors of the agents DMMP and GB quite differently. For Rehealing foams, the breakthrough times for GB are approximately 40% of those for DMMP, whereas for FC 602, the ratio is approximately 20%. CASCAD lies in between. This means that the foams do not act solely as physical barriers, but also have some chemical effects on the agents. For Rehealing and FC 602 with pH values of 6.5 and 6.0, respectively, this is quite surprising. Not so surprising CASCAD, with a pH of 10 and a good load of active chlorine reacts with GB: Whereas GB breaks through a 7 cm layer of CASCAD foam, it is chemically destroyed by a 10cm layer. DMMP on the other hand is more stable and breaks through the 10 cm of CASCAD.
For the life-size experiment, we measured the concentration of the simulant methyl salycilate (MS) in the air of the climate chamber over time after application of foam over the tub. [Figure 8] shows these concentrations for the five foams: CASCAD, Rehealing, FC 602, Jomos and MXOL, as well as for the uncovered tub.
Most importantly, there is no evidence that the simulant was spread out and the MS concentration at time 0 min did not increase at the application of foams. The vapor retaining effect appears to vary with the different foams; any foam though is better than the open tub, in which the MS concentration rises out of range of the Swiss CAM within less than a minute. The modulation of the curves is typical of the response of the Swiss CAM’s sensitivity to the oscillations of the humidity generator (±5 % r.h.).
In this paper, we presented our experimental approach to investigate the effect of foams on chemical agents in some detail. We found that it is not easy to perform reproducible experiments with foams. In spite of this, we were able to produce useful conclusions and clearly identify the suitability of the foams in destroying chemicals and suppressing the vapor above the surface to which the foams were applied.
The foam most effective for destroying chemical agents appears to be CASCAD. With a layer of seven cm, it prevented GB vapors from breaking through the foam for 200 min. Ten cm of CASCAD foam seemed to contain enough active chlorine to destroy 0.5 mL of GB and prevented any breakthrough. Breakthrough with GD (10 cm foam) occurred after 326 min, suggesting that some of the agent might have been destroyed. Breakthrough by HD occurred at 208 min, but there was no evidence of HD being destroyed by the active chlorine in the foam.
We also evaluated conventional fire fighter foams, the best ones are Jomos, Rehealing and FC 602, with GB breakthrough times of 235 min, 85 min and 63 min, respectively. Their pH values are between 6.0 and 6.5, and consequently these foams do not have much chemical activity (hydrolysis); these foams act mainly as physical barriers, although they react with the chemical agents to some extent.
For DMMP, the breakthrough times for CASCAD, Rehealing and FC 602 are 1095 min, 211 min and 324 min, respectively. These are approximately 2.5 to 5 times longer than the breakthrough times for GB, which corresponds with the lower vapor pressure of DMMP. The Jomos/DMMP combination could not be measured, because all chemical agent monitors were strongly cross sensitive to the foam, and the Hapsite Smart II showed interfering memory effects for DMMP which affected its ability to detect DMMP.
The remaining foams were not stable; Arctic, LS TS, MXOL, LW TN and Moussol had D90 s of 25min, 23 min, 15 min, 15 min and 8 min, respectively, and were not investigated any further. Based on the correlation of dehyration times with breakthrough times shown in [Figure 6], Arctic and LS TS were expected to have breakthrough times similar to FC 602. The others (MXOL, LW TN and Moussol) had even shorter lives and therefore were not tested further.
As for the thickness of the foam layer, the rule "the thicker the better" applies. We limited our experiments to 10 cm by considering practicality of the use of portable fire extinguishers and the amount of foam needed to cover an agent spill with such a layer.
From our “life-size” experiments, we found no evidence that the simulant was spread out by the application of foam. Obviously, the foam’s vapor retaining effect varies with the different foams; as a rule though, any foam is better than no foam, i.e., an uncovered source of vapors. This is of course valid only for chemicals with which a violent reaction is not expected.
We thank Christian Schäfer for his idea to monitor foam dehydration with sedimentation vessels, his miniature foam generator and two foam samples. Many thanks go also to Christophe Curty, Benjamin Menzi and Roland Kurzo from the organic synthesis group who supplied the chemical warfare agents, and Roland Liebi from personal NBC protection for lending us his AP2C measuring setup.
Figure 1: Experimental setup for breakthrough measurements. This is referred to as the “chemical measurement lid” in the text. From the mass flow controller, the air stream flushes the vapor space from three ports. The air from the exit port flows to a T-tube, from which the detector, in this case an Inficon Hapsite Smart II, draws air from one port for the analysis. Excess air escapes through the other port to vent. This arrangement yields constant dilution in the vapor space irrespective of monitor suction. [go to text reference]
Figure 2: Foam characteristic measurements. One liter graded conical glass sedimentation vessel used to monitor foam quality by measuring the dehydration rate of the foams. The characteristic parameter calculated from these measurements was the time to reach 90% dehydration, i.e., D90. See text for details. [go to text reference]
Figure 3: The effect of foam application on CWA spreading. Application of CASCAD foam from a backpack to a layer of 0.2 mm of methyl salicylate, MS, in a 60 cm x 60 cm x 20 cm tub from a distance of 3 m. Wind direction is left to right. The Swiss CAM is positioned upwind and measures the average MS concentration in the climate chamber. [go to text reference]
Figure 4: Concentration of phosphorus in the vapor space above the foam as a function of time for GB and CASCAD foam with foam layers of 3, 5, 7 and 10 cm. The different colors correspond to different data files. 22 µg/m3 of phosphorus correspond to 100 µg/m3 of GB, the nominal breakthrough concentration. [go to text reference]
Figure 5: Breakthrough times as a function of foam layer thickness. For the four longest lasting foams, DMMP and GB have been run. GD and HD have been measured for CASCAD, only. Two experiments, under identical conditions, were done for each foam and agent combination, designated as the diamonds and squares. For technical reasons, no DMMP could be run for Jomos foam. GB did not break through CASCAD foam at 10 cm thickness, i.e. these breakthrough times would be at infinity. [go to text reference]
Figure 6: Correlation of dehydration times D90 and breakthrough times for GB for a choice of foams for a layer of 5 cm. [go to text reference]
Figure 7: Correlation of breakthrough times with foam thickness. Using the same thickness of foams, the breakthrough times were measured in a series of layers of CASCAD, FC 602 and Rehealing foam. For example, a given data point correlates the breakthrough times of GB and DMMP through a 5 cm layer of CASCAD foam. The slopes of the linear regression lines are an illustration of how differently GB and DMMP vapors are confined by a given foam. A slope of 45° would mean that GB and DMMP vapors are confined equally well. The scatter of the CASCAD data is significantly larger than that of the other foams because of experimental difficulties adjusting the pressure during the production of the foam. (The slopes are y=0.391x for Rehealing, y=0.271x for CASCAD, and y=0.21x for FC 602.). [go to text reference]
Figure 8: Concentration of the simulant MS in the air of the climate chamber as a function of time for the 5 foams CASCAD, Reheal, FC 602, Jomos and MXOL as well as for the uncovered tub. The foam thicknesses were 10, 8, 9, 10 and 3 cm, respectively. CASCAD and MXOL appear to destroy the simulant MS. [go to text reference]
Allen-Vanguard (2009), http.//www.allen-vanguard.com/ViewProduct.aspx?ProductID=721 (accessed 14July2009)
Aue, W.P. and Guidetti, F. (2003) Live agent decontamination field trials with HD, THD, VX, GD and TGD: Comparison of three decontamination chemicals on seven materials, SPIEZ LABORATORY report LS 2003-14
U.S. Army SBCCOM, Chemical Command (2003) Risk assessment of using firefighter protective ensemble with self-contained breathing apparatus for rescue operations during a terrorist chemical agent incident, http://www.mabas.org/PDF-Files/ffpe_scba_rescue_ops.pdf (accessed 15July2009)