- Email: [email protected]
Improved research-scale foam generator design and performance characterization
Journal of Loss Prevention in the Process Industries 39 (2016) 173e180
Contents lists available at ScienceDirect
Journal of Loss Prevention in the Process Industries journal homepage: www.elsevier.com/locate/jlp
Improved research-scale foam generator design and performance characterization Brian Harding 1, Bin Zhang 1, Yi Liu, Hao Chen, M. Sam Mannan* Mary Kay O'Connor Process Safety Center, Artie McFerrin Department of Chemical Engineering, Texas A&M University System, College Station, TX 778433122, USA
a r t i c l e i n f o
a b s t r a c t
Article history: Received 8 May 2015 Received in revised form 19 November 2015 Accepted 19 November 2015 Available online 26 November 2015
The release of a cryogenic, flammable liquid, such as LNG, poses a threat to individuals in the area of the release as well as responders who attempt to limit the damage of the release. The most common mitigation technique is high-expansion foam which can be used to blanket the liquid, reducing the accumulation of flammable vapor above the pool through a number of different mechanisms. Despite the effectiveness of high-expansion foam blanketing, there are many aspects of the interaction between foam and LNG that are unknown. A lab-scale high-expansion foam generator has been developed to allow the study of those interactions. Additionally, the novel foam generator design addresses many of the drawbacks of industrial-scale foam generators and allows researchers better control of the foam, while producing foam at rates that are conducive to lab applications. Foam was produced using the generator and expansion ratio and foam stability were measured to determine the quality. The generator was able to produce foam with expansion ratio between 298 and 892 that collapsed at an average rate of 0.4 cm per minute. This quality of the foam is comparable to industrial-scale foam generators and the foam production rate is between 1.2 and 2.2 m3/min, which fits lab-scale needs. The foam generator can also be used with other types of non-firefighting foam, such as decontamination foam for chemical, biological, or nuclear decontamination. © 2015 Elsevier Ltd. All rights reserved.
Keywords: High-expansion foam Foam generation Foam solution age Expansion ratio Foam stability
1. Introduction 1.1. Foam techniques and applications in the process industry One of the original uses for foam in the process industry is in fire suppression. Depending on the fuel, the type of foam used can differ significantly (Martin, 2012; Sthamer, 2012), however the basics of the foam are the same, with the main components being a surfactant, water, and air. Foams can be used to fight class A and B fires (Martin, 2012), however the type of foam depends on the fuel source, and foams effective against one fire class may not be effective against the other (Martin, 2012). A more recent application of foams is for decontamination. In this application, the foam has additional chemicals such as peroxides and chloride salts (Cronce, 2002), which can decontaminate dangerous substances, for example chemical and biological warfare
* Corresponding author. E-mail address: [email protected] (M.S. Mannan). 1 These authors contributed equally to this paper. http://dx.doi.org/10.1016/j.jlp.2015.11.016 0950-4230/© 2015 Elsevier Ltd. All rights reserved.
agents (Cronce, 2002) and industrial contaminants. Another application of expansion foam in the process industry is the hazards mitigation of LNG spills. Foam application for LNG spills can be used as either a preventative or protective measure. As a preventative measure, on one hand, the recent work reveals that the foam works by reducing the heat convection and radiation through the blanketing effect, thereby reducing the vaporization rate of the LNG pool(Zhang et al., 2015, 2014). On the other hand, as LNG vapors pass through the foam zone, they are heated by contact with the much warmer foam, reducing the density of LNG vapor and thus minimizing the size of the ignitable LNG vapor cloud (Hiltz, 1993). When the LNG pool has already been ignited, the foam can be used as a protectant as well, which works to suppress a fire by four major mechanisms. The foam smothers the fire, physically separates the flames from the fuel source, cools the applied objects, and reduces the ability for flammable vapors to come in contact with oxygen from the air (Chemguard, 2014a). In high expansion foam applications on cryogenic liquid fuels, the foam performs three of these tasks, but does not work to cool the fuel surface, as the
174
B. Harding et al. / Journal of Loss Prevention in the Process Industries 39 (2016) 173e180
temperature of the foam is much higher than that of the fuel; however, it does work to reduce the heat input through convection and back radiation of the flames (Zhang et al., 2014). Although high expansion foam generally cannot extinguish an LNG fire on its own, it suppresses the fire, and allows the fire fighter to approach the fire and apply other firefighting methods, e.g., dry chemical. When the fire is ultimately extinguished by other means, any foam that is still present can take on a preventative role, serving to prevent reignition (Martin, 2012). 1.2. Gaps of industrial generators As mentioned previously, foams have been used for many years, however there are still gaps in foam research in both the LNG field and decontamination field. One of the major gaps in LNG research is the understanding of the physical interaction between the foam and the LNG system in lab scale tests. In decontamination research, experimentation has been fairly minimal both due to the recency of the technology (Cronce, 2002; Tucker, 2008; Tucker et al., 2004) and the highly hazardous chemicals used as the contaminant in decontamination experiments (Love et al., 2011). In order to conduct these experiments, an improved foam generator, which will be explained in detail in this paper, was developed to meet the research demand. In a fire scenario, the application rate of the firefighting agent (i.e., water, dry chemical, or foam) is a crucial factor in extinguishing an existing fire or preventing a fire from spreading. For this reason, industrial foam generators are constructed to apply foam at a very high rate. Industrial foam generators, such as Chemguard 1500 WP foam generator, often are able to create foam at flow rates above 38 m3/min (Chemguard, 2014b), which is relatively low for an industrial foam generator (Angus Fire, 2014a, 2014b). Additionally, in order to increase the applicability of the foam generator, hydraulic power is typically used. Because fire codes require fire water to be pumped throughout the facility and accessible from fire hydrants at regular distance intervals (National Fire Protection Association [NFPA], 2012), the pressurized water is widely accessible in an industrial facility, which makes using hydraulic power an excellent approach. Although commercially available foam generators are suitable for foam application in industry, there are drawbacks of using them in a research setting. The most obvious problem is the foam application rate. During foam application on an industrial spill, high foam application rate is beneficial to cover the spill quickly; however, the same application rate is far too high for lab scale research. In an LNG spill scenario, characteristic foam depth for an LNG spill is anywhere from 0.45 m to 0.91 m (NFPA, 2013). Assuming a floor area of 55 m2, which is typical for a research lab, even at the minimum setting the foam from these commercially available foam generators would fill the lab to a depth of 0.5 m in 43 s. Additionally, industrial scale foam generators are powered by pressurized water, which causes two operational problems, the requirement for a large volume of pressurized water, and excessive water discharge that accompanies the introduction of water during startup. The dependence on pressurized water also poses the safety issue created by having a pressurized system. The pressurized water requirement limits the availability of such equipment only to areas where pressurized water is accessible. In certain applications water discharge during startup is tolerable; however, when applying the foam to cryogenic liquids, the excessive water discharge causes rapid vaporization of the liquid, which compromises the objective of foam application. Moreover, commercial foam generators provide little working flexibility aside from changing hydraulic pressure (Angus Fire, 2014a, 2014b, 2014c; Chemguard, 2014b). Some dependent variables such as foam application rate, foam expansion
ratio, and foam bubble size are important in research on foam functionality. The fact that these variables are inextricably related with others requires independent control of each parameter to study the effect of individual variable on foam functionality. Therefore, it is desired to manipulate those variables in an organized manner from the standpoint of experimental design. The ultimate purpose of this work is to provide a feasible design for a research scale foam generator and discuss several key parameters of foam functionality such as foam expansion ratio, time to halfheight, and foam application rate associated with such design. Utilizing the design proposed in this work will help disclose the effect of individual variables on foam performance for different applications in the future including LNG spill control, decontamination, and fire suppression. 2. Foam generator design and construction 2.1. Foam generator design The research scale foam generator was designed according to “NFPA 11 Standard for Low-, Medium-, and High-Expansion Foam” (NFPA, 2012). A picture of the conceptual device in NFPA 11 is shown in Fig. 1. The design also shares some similarities with aspects of previous foam generator patents, such as the position of the nozzle and screen and the use of a fan to generate airflow (Jamison and Barnes, 1965; Jamison, 1966; O'Regan et al., 1970; Williams, 1953). As mentioned previously, the main downsides of the industrial scale foam generator are: high foam application rate, dependence on hydraulic power, and lack of customizability. In order to address high foam application rate, the device was built on a much smaller scale with the goal of foam application rate being less than 2.8 m3/min (compared to the 38 m3/min minimum of industrial scale generators). Additionally, the device was constructed with an effort to minimize dependence on utilities, such as pressurized air and water. In the current setup, the only required
Fig. 1. High expansion foam quality test generator (NFPA, 2012).
B. Harding et al. / Journal of Loss Prevention in the Process Industries 39 (2016) 173e180
utility is electricity, which allows more portability and availability of the equipment. The independence of the equipment on pressurized water addresses both the availability and water discharge issues of the industrial equipment. Lack of customizability is addressed by building the generator from individual parts. Building a foam generator allows the researcher to decide which components are used for each element of the machine, and can decide to build a machine with much more customizability. The foam generating apparatus constructed in the lab allows the adjustment, either through substitution or control, of mesh size, nozzle position, nozzle type, air flow rate, foam solution pressure, foam solution composition, and foam application angle, giving the researcher many options to control the output of a single dependent variable. 2.2. Construction The original design for the apparatus provided in NFPA 11 shows the basics of how to construct such a device, but leaves a lot of the design decisions up to the researcher. As shown in Fig. 1, the foam solution is pushed through the pipes by backpressure from the air supply and dispersed through the spray nozzle onto the screen where air from the blower moves through the screen creating foam bubbles. The apparatus also contains a pressure gauge to monitor foam solution pressure, a pressure regulator to control the pressure of the air inlet, a metering valve to control the flow rate of the foam solution, and an adjustable damper to control the flow rate of the air. Foam solution flow rate through the nozzle is a function of pressure, therefore to ensure accurate flow rate data, the pressure readings should be taken as close to the nozzle as possible. Additionally, there is a solenoid valve for remote control, a bleed valve to depressurize the system during shutdown, and a liquid pressure gauge to take accurate pressure readings just before the nozzle. The transparency of the air cylinder is not a requirement, but could allow the nozzle and screen during operation to be visualized for better control. Many improvements have been performed based on the original model illustrated in NFPA11 for research purpose, such as the orientation of the transparent air cylinder, the inclusion of a pump, and the differences in the piping setup. A summary of major improvements is listed in Table 1. The new setup is shown in Fig. 2. The horizontal orientation of the transparent air cylinder performs two functions. The first is that a horizontal setup allows foam to be carried by the air flow toward a target. This adjustment allows the flexibility of developing foam that is deposited into a foam collection tank directly beneath the foam solution or projected toward a location 3e5 m away from the source. In order to direct the flow of foam, a reflector plate was installed at the end of the transparent air cylinder. The position of this plate can be adjusted to push the foam down at a variety of angles allowing the foam to be directed onto spills at various distances from the generator. The other function this orientation provides is the inclusion of the drip
175
tray. In the vertical setup, any foam solution that is not converted into foam will drip into the foam collection tank. In the new design, the excess foam solution settles in the drip tray which is positioned just after the screen. From there the foam solution moves through a hose back to the foam solution tank. The drip tray not only limits foam solution waste, but also helps in accurate characterization of expansion ratio and the effect of foam application on the vaporization rate of an LNG spill. Because the expansion ratio calculation is done by dividing the foam volume by the mass of foam solution used, any foam solution not converted to foam will increase the mass without increasing the foam volume, giving inaccurate data. When applied to cryogenic liquids, unconverted foam solution can also have detrimental effects, similar to the water discharge phenomenon. If unconverted foam solution is applied to an LNG pool, the large temperature difference between the foam solution and the cryogenic liquid pool and the high heat capacity of foam solution will cause additional LNG to evaporate, which is counterproductive to the application of foam and should be avoided in the study. The inclusion of a pump in the system is both a necessity and an improvement. During pressure testing, the foam solution tank, a 50 L LDPE carboy from US Plastic Corp, was found to withstand around 28 psig. However, the system pressure must be able to reach 50 psig to achieve the desired foam solution flow rate. To resolve this problem, the air supply was removed and a pump was added downstream of the foam solution tank to pressurize the foam solution. This new design also has some advantages over the previous design in NFPA 11. The first is that the new system only requires power to operate, meaning that it can be transported and used in areas that do not have a pressurized air source. The second improvement is that the pump can boost the pressure much higher than the pressurized air, allowing a higher maximum foam solution flow rate. Because a majority of the setup is made from stainless steel and the PVC components can withstand 160 psi, running at higher pressures is a viable option. The third improvement is that the new design is inherently safer and easier to operate. Because the pump is placed downstream of the foam solution tank, the pressurized section of the setup is smaller due to the exclusion of the tank. This modification makes the setup inherently safer due to a smaller pressurized volume. Furthermore, the removal of the pressurized air simplifies the startup and shutdown procedures, making operation of the equipment easier on the researcher. These two changes illustrate two of the principles of inherently safer design, minimize and simplify. The last improvement is that the drain can feed directly back into the unpressurized foam solution tank to minimize the foam solution losses, while in the original design, the drain would have to go to waste because it could not feed back into the high pressure system. The last major change is the difference in piping setup. In the original NFPA diagram, there is very little detail on how to actually assemble the apparatus. Many of the piping decisions were made
Table 1 Improvements of foam generator design. Original design in NFPA 11
New design in this work
Improvements
Vertical air cylinder
Horizontal air cylinder
Solution propelled by pressurized air Bleed valve used to release pressure during shutdown Solenoid valve used for remote startup and shutdown Vertical flow
Solution propelled by jet pump
Allows higher foam outlet point due to ceiling height restriction Improves safety of design and removes dependence on pressurized air Improves design safety and simplifies shutdown procedure With smaller pressurized volume, valve is no longer needed, startup and shutdown simplified Allows foam to be directed into foam collection tank for an even fill
Smaller pressurized volume (50 L in the original design to 1.5 L in the improved design), pressure naturally bleeds out the nozzle Removed Solenoid valve Inclusion of deflector plate or a plastic duct
176
B. Harding et al. / Journal of Loss Prevention in the Process Industries 39 (2016) 173e180
Closeup of Damper Damper Handle 100 % Open 50 % Open Unistrut® Frame (8' x 3')
10 % Open
Transparent Air Cylinder
Deflector Plate Nozzle
Airflow In
Elbow
Air Out
Nipple Screen Fan Adjustable Damper Foam SoluƟon Tank
Pressure Gauge
Foam Out
½” PVC Tubing
Ball Valve
Closeup of Screen Assembly Gasket Washer Washer Gasket
Tap
½” Stainless Tubing
¾” Stainless Tubing
Bolt Screen Nut Flange Flange
Pump Pressure Regulator
Fig. 2. Diagram of final high expansion foam generator.
during the planning and construction of the improved apparatus, some of which add to the functionality of the setup, and others which are simply required for the foam solution to flow through the system. One such addition is the 10 foot flexible PTFE/stainless steel hose. This addition allows the height of the air cylinder to be changed, permitting the application of foam into containers of different height or the application of foam at different trajectories. Another addition is the elbow/nipple combination located before the nozzle. This section of the piping allows nozzle repositioning in reference to the screen, ensuring that the screen is properly coated with foam solution and no solution is sprayed on the sides of the air cylinder. There are also multiple nozzles and multiple screens with different mesh sizes that can be selected depending on the application. The transparency of the air cylinder is not a requirement, but could allow the nozzle and screen to be visible during operation for better control. In order for the apparatus to be effective, it must be able to continuously apply high quality foam at the desired expansion ratio. In order to achieve this goal, many of the aspects of the apparatus must be adjustable to reach the desired foam application. The aspects of the foam that can be controlled are the foam flow rate, foam expansion ratio, and foam bubble size. Each of these aspects is controlled by adjusting multiple different components of the system. The foam flow rate can be changed by changing the nozzle
from a low flow nozzle to high flow nozzle and by increasing the pressure by turning up the pump and the pressure regulator. The foam expansion ratio can be changed by changing the foam solution flow rate as mentioned before in combination with the air flow rate from the fan. The foam bubble size can be changed by adjusting the mesh size of the screen and the air flow rate. 3. Characterization and discussion In order to test the validity of the foam generating apparatus, foam was produced and characterized. The three important characteristics of the foam are expansion ratio, time to half-height, and foam application rate. Expansion ratio is defined as the ratio of the amount of air to the amount of liquid foam solution in a foam. In firefighting, foams with an expansion ratio less than 20:1 are called low-expansion foams, between 20:1 and 200:1 are called mediumexpansion foams, and greater than 200:1 are called high expansion foams (NFPA, 2012). Each type of foam has advantages and disadvantages depending on the application scenarios. Low expansion foams can behave like low-viscosity liquids, which make them ideal for covering large areas quickly (NFPA, 2012). Also, due to the lower air content, and associated higher specific gravity, they are affected less by the wind and other adverse weather conditions (NFPA, 2012). Additionally, they can be blasted or thrown similar to water
B. Harding et al. / Journal of Loss Prevention in the Process Industries 39 (2016) 173e180
coming out of a fire hose (Martin, 2012). Because their density is much lower than that of water, they have much less inertia and therefore cannot be blasted as far, but they can still be used on fires that are far away. On the other end of the spectrum, high expansion foams are more useful when trying to fill a space, because of the high expansion ratio (Martin, 2012). If, for instance, there is a fire in a storage facility or a hangar, the foam can be used to quickly fill the structure without using large quantities of foam solution (Martin, 2012). High expansion foams are also ideal for use on LNG or other cryogenic liquid spills. The low liquid content of the foam contributes less latent heat to the spill while still providing the heat insulation and physical barrier effects. High expansion foams cannot be blasted because of their very low density. These foams must be produced and deposited directly on the area of application. Medium expansion foams have some of the applicability of each of the previous types. They can be blasted about 12e35 m, depending on the expansion ratio, making them very versatile (Sthamer, 2012). 3.1. Expansion ratio As mentioned before, expansion ratio is defined as the ratio of the amount of air to the amount of liquid foam solution in a foam. The two contributing factors to this ratio are the air flow rate and the liquid flow rate. The air flow rate is controlled by the damper, positioned behind the fan. In this study, the damper conditions of 100% open, 50% open, 25% open, and 12.5% open, as well as liquid flow rates between 1.4 and 5.7 L/min, were used. The liquid flow rate is controlled by the pressure regulator positioned directly after the pump. The foam collection tank is positioned on a balance, which is used to measure the mass change in the tank. With a constant liquid flow rate, increasing the air flow rate will cause the expansion ratio to increase, assuming the air is being trapped within the bubbles. With the mass change and the total volume of the foam collection tank, the expansion ratio can be determined. There is a very slight mass contribution from the air trapped in the foam bubble due to the increased pressure in the bubble, however this contribution is approximately 2.8 g, which represents a minimal effect and can be neglected. From Fig. 3, it can be found that all the foams have expansion ratio higher than 200, indicating they all are high expansion foams, and the foam expansion ratio decreases with increasing the liquid flow rate (affected by applied pressure) when maintaining constant
177
aperture size. Although only high expansion foams were created in these tests, using different conditions and a different foam solution could yield foams with much lower expansion ratios, if a lower expansion ratio was desired. An interesting result was the similarity of the 100% open and 25% open results. A possible explanation is that with the damper 100% open, the air flow is in excess of what it takes to create foam, and for this reason the foam being created in both tests is very similar. When running the test with aperture 100% open, it was observed that a considerable amount of air is not entrained in the bubbles, instead, it is used to carry the foam giving it a much straighter and more forceful horizontal trajectory. However, in the 25% open test, the foam follows a parabolic trajectory and easily flows downward into the foam collection tank. This difference in trajectory is due to the difference in air flow. Another interesting phenomenon is the higher expansion ratio of the 50% open damper position at low foam solution flow rate. Where 25% open and 100% open begin to flatten out, 50% continues its upward trend. This is probably due to multiple factors affecting the bubble formation. In the 50% open trials at 2.3 L/min and 3 L/ min, although the expansion ratio was higher than any other trials with the large nozzle, the foam stability was much lower than trials with a similar foam solution flow rate, indicating that something about the foam bubbles may have been fundamentally different, such as the thickness of the bubble wall. The bulk of the trials were run using the WL 1½ nozzle from BETE, which produces a flow rate of 5.7 L/min at 40 psi (BETE, 2014). In order to test even lower flow rates, which were expected to have the highest expansion ratios, a smaller nozzle, called the WL 1, was used. This nozzle produces a flow rate of 3.8 L/min at 40 psi. Trials to the left of the vertical line in Fig. 3 are using this smaller nozzle. Through the use of the smaller nozzle, an expansion ratio of 852 was achieved, which is among the highest in any of the trials. As the pressure drops below about 5 psi, the conical spray pattern of the nozzle narrows into a jet and does a poor job of covering the screen. Switching to a smaller nozzle allows the researcher to achieve a lower foam solution flow rate while maintaining a pump pressure above 5 psi. Trials were also conducted using a pressure of 11.5 psi with the WL 1 nozzle, which gives an equivalent foam solution flow rate to 5 psi using the WL 1½ nozzle. In all of these trials the expansion ratio was higher using the small nozzle than the equivalent flow rate using the larger nozzle, which can be seen in Fig. 4. The likely explanation is that by operating at a higher pressure, the spray pattern is able to cover a larger area of the screen and do so with more uniformity, allowing additional air to be entrained in the bubbles creating a higher expansion ratio foam. 3.2. Time to half-height
Fig. 3. Expansion Ratio vs. Foam Solution Flow Rate for all expansion foam trials. Values in the legend represent aperture position and are written as percentage of the area open to flow. Trials left of the vertical line were conducted using the smaller nozzle, all others were conducted using the larger nozzle.
The term “metric quarter time” is usually used to determine the rate of liquid drainage from the foam matrix (Conroy et al., 2013). Quarter time is defined as the time it takes one quarter of the foam solution to drain out of the foam matrix. In the current setup, draining rate cannot be measured, however the height of the foam in the container can be measured accurately. For this reason, the stability metric will be time to half-height, which is defined as the time the foam takes to settle to half of its original height. The foam addition and subsequent collapse has been captured by video camera for each experiment and analyzed to determine time to half-height. Each trial was filmed in order to collect real time height information for the collapse rate. The foam collection tank was marked with horizontal lines every six inches to provide a scale on the camera. The height at the center was used as the reference point (18 inches from each wall). Time to half-height is used as the main
178
B. Harding et al. / Journal of Loss Prevention in the Process Industries 39 (2016) 173e180
Fig. 4. Comparison of expansion ratio at different damper positions using the Small Nozzle (WL1) and Big Nozzle (WL1½) keeping foam solution flow rate constant.
comparison tool, however for certain trials with interesting characteristics, MATLAB was used to obtain height information on 1-s intervals and observe the entire foam height curve. For foam stability, assuming similar bubble size, foams with higher expansion ratios should have a shorter time to half-height, because there is less foam solution present in the initial foam. In this data the trend is present, but the correlation is very weak. In Fig. 5, two data points are highlighted, Trial A and Trial B. Both are run at identical foam generator conditions (Pump pressure 30 psi, damper position 25% open) and have similar expansion ratios (Trial A was 405, Trial B was 393), however their stability characteristics are dramatically different. The Trial A time to half-height was 282 min and the Trial B time to half-height was 151 min, representing a decrease of more than 46%. In Trial A, the foam solution had been mixed approximately 28 h before application, whereas in Trial B, the foam solution had been mixed approximately 220 h before application, giving the foam solution in Trial B much more time to suffer stability losses. The potential causes for this drop in stability are the age of the foam solution, the interaction of the foam solution with the pipe work, and the interaction of the foam
Fig. 5. Time to Half-Height vs. Expansion Ratio for all large nozzle trials. Trials A and B have been highlighted to show the effect of foam solution age on time to half-height.
Fig. 6. Time to Half-Height vs. Foam Solution Age for exposed and unexposed scenarios.
solution with air. These effects on foam stability prompted a series of experiments on measuring foam stability with foam solution age as the independent variable. The results of these experiments are shown in Fig. 6. In the first study, which focuses on the effect of foam solution age, the foam solution was kept in a sealed carboy to limit the amount of exposure to the air. These experiments show no decrease in time to half-height with increased age, meaning that the drop in foam stability was likely due to interaction with either air or the pipework. The second study was conducted with the foam solution left in the foam solution tank, which is not isolated from the air, and a significant drop in time to half-height occurred with foam solution older than eight days. This study also served to eliminate the potential factor of foam solution interaction with the pipe work due to the offset time in between trials. The first trial (at one day) was run immediately, with no time spent in the pipework and no time spent exposed to air. The second trial (at eight days) was run with seven days spent exposed to the pipework and seven days spent exposed to the air. The third trial (at 10 days) was run with two days spent exposed to the pipework and nine days spent exposed to the air. If pipework exposure was the determining factor in foam stability, then the time to half-height for the trial at 10 days
Fig. 7. Foam breaking info for a full trial with a linear trend line for visual reference.
B. Harding et al. / Journal of Loss Prevention in the Process Industries 39 (2016) 173e180
179
4. Conclusions 4.1. Main features and the benefits on foam related research
Fig. 8. Foam Production Rate (m3/min) for Expansion Foam Tests. Only trials using the larger nozzle are depicted. Values in the legend represent aperture position and are written as percentage of the area open to flow.
would be expected to be higher than the time to half-height at eight days. From these trials, the conclusion can be made that exposure to air is the main contributor to a decrease in foam solution stability. Fig. 7 shows the foam height throughout a single trial to illustrate how foam height changes with time. In this trial the pressure was 5 psi and the damper was set at 25% open, resulting in an expansion ratio of 643 and a time to half-height of 244 min. For the purposes of comparison, the linear fit predicts a time to half-height of 245.4 min. For this analysis, the video was broken into pictures at 1 min intervals and MATLAB was used to analyze the position of the upper surface of the foam. From that information, foam height could be calculated accurately across the trial. As the figure demonstrates, there is a highly linear dependence of height on time, with an R2 value of 0.978. This linearity is also present in other trials regardless of application conditions.
3.3. Foam application rate In addition to foam quality characterization such as expansion ratio and time to half-height, another important criterion of foam generation is foam application rate. The application time is the period between the moment the first bubble of foam entered the foam collection tank to the moment the foam generator was unpowered, recorded by a video camera. The foam application rate is determined by dividing the total volume of the foam collection tank by application time, the error of which is estimated to be within 4% due to a trivial part of foam spill out of the tank. Foam application rates for all of the trials were found to be between 1.2 and 2.2 m3/min, which are well suited for lab-scale research purpose, compared with the high application rates of commercial generators (larger than 38 m3/min). As shown in Fig. 8, the foam production rate for the damper position at 12.5% open was much lower than the production rate at other damper settings because the air flow rate was the limiting factor in foam formation. The foam production rate is essentially a measure of how much air is trapped in foam bubbles and when the air flow rate drops below a threshold value, the foam production rate drops along with it. This same effect can be seen in the expansion ratio data in Fig. 3, where the limiting air flow rate caused a decrease in expansion ratio when compared to other damper settings.
The foam generating apparatus discussed in this paper was able to address all of the issues that exist with using an industrial scale foam generator for lab use. The foam generator was able to produce high quality foams at flow rates that fit lab scale research. These foams had expansion ratios between 298 and 892, with an average time to half-height around 185 min. The uncoupling of the air flow rate and the liquid flow rate also allows the researcher customizability without changing components of the generator, which is an advantage over industrial scale foam generators. The portability and availability of the equipment, lack of water discharge at the beginning of a trial, and lower cost of building equipment are also advantages to the researcher that this equipment can offer. The constructed generator also exhibited good customizability, allowing the researcher to substitute many of the parts in order to create a foam with a variety of different physical characteristics. The main variables that were adjusted were foam solution pressure, damper position, and nozzle size, each of which showed a strong effect on the resulting foam. In addition to these variables, many other adjustments can be made to produce a desirable, applicable foam. The main improvements as they pertain to LNG research are the control over independent variables and the lack of water discharge at startup. With these improvements over the commercial foam generator, researchers will be better able to conduct trials with specific variables in mind and with a high control over the conditions present in the experiment, allowing a more in depth study of the factors related with foam application to LNG spills or other foam related research. References Angus Fire, 2014a. Fixed Turbex Systems (WWW Document). http://www.angusfire. co.uk/products/foam_equipment/fixed/turbexlng.html (accessed 6.13.14). Angus Fire, 2014b. LNG Turbex Skids (WWW Document). http://www.angusfire.co. uk/products/foam_equipment/fixed/lngturbexskids.html (accessed 6.13.14). Angus Fire, 2014c. Mini-turbex (WWW Document). http://www.angusfire.co.uk/ products/foam_equipment/portable/miniturbex.html (accessed 6.13.14). BETE, 2014. Open Performance Data Table - SI/Metric Units (WWW Document). http://www.bete.com/pdfs/BETE_WL-metric.pdf (accessed 2.8.14). Chemguard, 2014a. General Foam Information (WWW Document). http://www. chemguard.com/about-us/documents-library/foam-info/general.htm (accessed 4.2.14). Chemguard, 2014b. Water Powered High Expansion Foam Generator (WWW Document). http://www.chemguard.com/fire-suppression/catalog/EngineeredSystems/high-expansion-foam-generators/1500wp.aspx (accessed 6.13.14). Conroy, M.W., Taylor, J.C., Farley, J.P., Fleming, J.W., Ananth, R., 2013. Liquid drainage from high-expansion (HiEx) aqueous foams during and after filling of a container. Colloids Surf. A Physicochem. Eng. Asp. 426, 70e97. http://dx.doi.org/ 10.1016/j.colsurfa.2013.02.050. Cronce, D.T., 2002. Chemical Warfare Agent Decontamination Foaming Composition and Method, 6376436 B1. Hiltz, R., 1993. Foam Blanketing: The Use of Foam to Mitigate the Vapor Hazard of Spilled Volatile Chemicals. In: Prevention and Control of Accidental Releases of Hazardous Gases, pp. 216e231. Jamison, W.B., 1966. Fire-fighting Foam Generator, 3241617. Jamison, W.B., Barnes, R.W., 1965. Fire-fighting Method Employing High Expansion Foam, 3186490. Love, A.H., Bailey, C.G., Hanna, M.L., Hok, S., Vu, A.K., Reutter, D.J., Raber, E., 2011. Efficacy of liquid and foam decontamination technologies for chemical warfare agents on indoor surfaces. J. Hazard. Mater 196, 115e122. http://dx.doi.org/ 10.1016/j.jhazmat.2011.09.005. Martin, T.J., 2012. In: Stevenson, P. (Ed.), Fire-fighting Foam Technology. John Wiles & Sons, Ltd, Chichester, UK, pp. 411e457. NFPA 11: Standard for Low-, Medium-, and High-Expansion Foam, 2012. NFPA 24: Standard for the Installation of Private Fire Service Mains and Their Appurtenances, 2013. O'Regan, J.F., Lundberg, B.A., Mussoni, W.J., 1970. High Expansion Foam Generator, 3512761. Sthamer, R., 2012. Medium Expansion Foam (WWW Document). http://sthamer. com/englisch/f32_medium_foam.html# (accessed 10.21.13).
180
B. Harding et al. / Journal of Loss Prevention in the Process Industries 39 (2016) 173e180
Tucker, M.D., 2008. Enhanced Formulations for Neutralization of Chemical Biological and Industrial Toxants, 7390432 B2. Tucker, M.D., Betty, R.G., Tadros, M.E., 2004. Concentrated Formulations and Methods for Neutralizing Chemical and Biological Toxants, 6723890 B2. Williams, E.C., 1953. Apparatus for Extinguishing Fires, 2645292. Zhang, B., Harding, B., Liu, Y., Mannan, M.S., 2015. Mitigation effect of high
expansion foam on LNG vapor hazard. In: 11th Global Congress on Process Safety. Austin, Texas. Zhang, B., Liu, Y., Olewski, T., Vechot, L., Mannan, M.S., 2014. Blanketing effect of expansion foam on Liquefied Natural Gas (LNG) spillage pool. J. Hazard. Mater 280, 380e388. http://dx.doi.org/10.1016/j.jhazmat.2014.07.078.
Contents lists available at ScienceDirect
Journal of Loss Prevention in the Process Industries journal homepage: www.elsevier.com/locate/jlp
Improved research-scale foam generator design and performance characterization Brian Harding 1, Bin Zhang 1, Yi Liu, Hao Chen, M. Sam Mannan* Mary Kay O'Connor Process Safety Center, Artie McFerrin Department of Chemical Engineering, Texas A&M University System, College Station, TX 778433122, USA
a r t i c l e i n f o
a b s t r a c t
Article history: Received 8 May 2015 Received in revised form 19 November 2015 Accepted 19 November 2015 Available online 26 November 2015
The release of a cryogenic, flammable liquid, such as LNG, poses a threat to individuals in the area of the release as well as responders who attempt to limit the damage of the release. The most common mitigation technique is high-expansion foam which can be used to blanket the liquid, reducing the accumulation of flammable vapor above the pool through a number of different mechanisms. Despite the effectiveness of high-expansion foam blanketing, there are many aspects of the interaction between foam and LNG that are unknown. A lab-scale high-expansion foam generator has been developed to allow the study of those interactions. Additionally, the novel foam generator design addresses many of the drawbacks of industrial-scale foam generators and allows researchers better control of the foam, while producing foam at rates that are conducive to lab applications. Foam was produced using the generator and expansion ratio and foam stability were measured to determine the quality. The generator was able to produce foam with expansion ratio between 298 and 892 that collapsed at an average rate of 0.4 cm per minute. This quality of the foam is comparable to industrial-scale foam generators and the foam production rate is between 1.2 and 2.2 m3/min, which fits lab-scale needs. The foam generator can also be used with other types of non-firefighting foam, such as decontamination foam for chemical, biological, or nuclear decontamination. © 2015 Elsevier Ltd. All rights reserved.
Keywords: High-expansion foam Foam generation Foam solution age Expansion ratio Foam stability
1. Introduction 1.1. Foam techniques and applications in the process industry One of the original uses for foam in the process industry is in fire suppression. Depending on the fuel, the type of foam used can differ significantly (Martin, 2012; Sthamer, 2012), however the basics of the foam are the same, with the main components being a surfactant, water, and air. Foams can be used to fight class A and B fires (Martin, 2012), however the type of foam depends on the fuel source, and foams effective against one fire class may not be effective against the other (Martin, 2012). A more recent application of foams is for decontamination. In this application, the foam has additional chemicals such as peroxides and chloride salts (Cronce, 2002), which can decontaminate dangerous substances, for example chemical and biological warfare
* Corresponding author. E-mail address: [email protected] (M.S. Mannan). 1 These authors contributed equally to this paper. http://dx.doi.org/10.1016/j.jlp.2015.11.016 0950-4230/© 2015 Elsevier Ltd. All rights reserved.
agents (Cronce, 2002) and industrial contaminants. Another application of expansion foam in the process industry is the hazards mitigation of LNG spills. Foam application for LNG spills can be used as either a preventative or protective measure. As a preventative measure, on one hand, the recent work reveals that the foam works by reducing the heat convection and radiation through the blanketing effect, thereby reducing the vaporization rate of the LNG pool(Zhang et al., 2015, 2014). On the other hand, as LNG vapors pass through the foam zone, they are heated by contact with the much warmer foam, reducing the density of LNG vapor and thus minimizing the size of the ignitable LNG vapor cloud (Hiltz, 1993). When the LNG pool has already been ignited, the foam can be used as a protectant as well, which works to suppress a fire by four major mechanisms. The foam smothers the fire, physically separates the flames from the fuel source, cools the applied objects, and reduces the ability for flammable vapors to come in contact with oxygen from the air (Chemguard, 2014a). In high expansion foam applications on cryogenic liquid fuels, the foam performs three of these tasks, but does not work to cool the fuel surface, as the
174
B. Harding et al. / Journal of Loss Prevention in the Process Industries 39 (2016) 173e180
temperature of the foam is much higher than that of the fuel; however, it does work to reduce the heat input through convection and back radiation of the flames (Zhang et al., 2014). Although high expansion foam generally cannot extinguish an LNG fire on its own, it suppresses the fire, and allows the fire fighter to approach the fire and apply other firefighting methods, e.g., dry chemical. When the fire is ultimately extinguished by other means, any foam that is still present can take on a preventative role, serving to prevent reignition (Martin, 2012). 1.2. Gaps of industrial generators As mentioned previously, foams have been used for many years, however there are still gaps in foam research in both the LNG field and decontamination field. One of the major gaps in LNG research is the understanding of the physical interaction between the foam and the LNG system in lab scale tests. In decontamination research, experimentation has been fairly minimal both due to the recency of the technology (Cronce, 2002; Tucker, 2008; Tucker et al., 2004) and the highly hazardous chemicals used as the contaminant in decontamination experiments (Love et al., 2011). In order to conduct these experiments, an improved foam generator, which will be explained in detail in this paper, was developed to meet the research demand. In a fire scenario, the application rate of the firefighting agent (i.e., water, dry chemical, or foam) is a crucial factor in extinguishing an existing fire or preventing a fire from spreading. For this reason, industrial foam generators are constructed to apply foam at a very high rate. Industrial foam generators, such as Chemguard 1500 WP foam generator, often are able to create foam at flow rates above 38 m3/min (Chemguard, 2014b), which is relatively low for an industrial foam generator (Angus Fire, 2014a, 2014b). Additionally, in order to increase the applicability of the foam generator, hydraulic power is typically used. Because fire codes require fire water to be pumped throughout the facility and accessible from fire hydrants at regular distance intervals (National Fire Protection Association [NFPA], 2012), the pressurized water is widely accessible in an industrial facility, which makes using hydraulic power an excellent approach. Although commercially available foam generators are suitable for foam application in industry, there are drawbacks of using them in a research setting. The most obvious problem is the foam application rate. During foam application on an industrial spill, high foam application rate is beneficial to cover the spill quickly; however, the same application rate is far too high for lab scale research. In an LNG spill scenario, characteristic foam depth for an LNG spill is anywhere from 0.45 m to 0.91 m (NFPA, 2013). Assuming a floor area of 55 m2, which is typical for a research lab, even at the minimum setting the foam from these commercially available foam generators would fill the lab to a depth of 0.5 m in 43 s. Additionally, industrial scale foam generators are powered by pressurized water, which causes two operational problems, the requirement for a large volume of pressurized water, and excessive water discharge that accompanies the introduction of water during startup. The dependence on pressurized water also poses the safety issue created by having a pressurized system. The pressurized water requirement limits the availability of such equipment only to areas where pressurized water is accessible. In certain applications water discharge during startup is tolerable; however, when applying the foam to cryogenic liquids, the excessive water discharge causes rapid vaporization of the liquid, which compromises the objective of foam application. Moreover, commercial foam generators provide little working flexibility aside from changing hydraulic pressure (Angus Fire, 2014a, 2014b, 2014c; Chemguard, 2014b). Some dependent variables such as foam application rate, foam expansion
ratio, and foam bubble size are important in research on foam functionality. The fact that these variables are inextricably related with others requires independent control of each parameter to study the effect of individual variable on foam functionality. Therefore, it is desired to manipulate those variables in an organized manner from the standpoint of experimental design. The ultimate purpose of this work is to provide a feasible design for a research scale foam generator and discuss several key parameters of foam functionality such as foam expansion ratio, time to halfheight, and foam application rate associated with such design. Utilizing the design proposed in this work will help disclose the effect of individual variables on foam performance for different applications in the future including LNG spill control, decontamination, and fire suppression. 2. Foam generator design and construction 2.1. Foam generator design The research scale foam generator was designed according to “NFPA 11 Standard for Low-, Medium-, and High-Expansion Foam” (NFPA, 2012). A picture of the conceptual device in NFPA 11 is shown in Fig. 1. The design also shares some similarities with aspects of previous foam generator patents, such as the position of the nozzle and screen and the use of a fan to generate airflow (Jamison and Barnes, 1965; Jamison, 1966; O'Regan et al., 1970; Williams, 1953). As mentioned previously, the main downsides of the industrial scale foam generator are: high foam application rate, dependence on hydraulic power, and lack of customizability. In order to address high foam application rate, the device was built on a much smaller scale with the goal of foam application rate being less than 2.8 m3/min (compared to the 38 m3/min minimum of industrial scale generators). Additionally, the device was constructed with an effort to minimize dependence on utilities, such as pressurized air and water. In the current setup, the only required
Fig. 1. High expansion foam quality test generator (NFPA, 2012).
B. Harding et al. / Journal of Loss Prevention in the Process Industries 39 (2016) 173e180
utility is electricity, which allows more portability and availability of the equipment. The independence of the equipment on pressurized water addresses both the availability and water discharge issues of the industrial equipment. Lack of customizability is addressed by building the generator from individual parts. Building a foam generator allows the researcher to decide which components are used for each element of the machine, and can decide to build a machine with much more customizability. The foam generating apparatus constructed in the lab allows the adjustment, either through substitution or control, of mesh size, nozzle position, nozzle type, air flow rate, foam solution pressure, foam solution composition, and foam application angle, giving the researcher many options to control the output of a single dependent variable. 2.2. Construction The original design for the apparatus provided in NFPA 11 shows the basics of how to construct such a device, but leaves a lot of the design decisions up to the researcher. As shown in Fig. 1, the foam solution is pushed through the pipes by backpressure from the air supply and dispersed through the spray nozzle onto the screen where air from the blower moves through the screen creating foam bubbles. The apparatus also contains a pressure gauge to monitor foam solution pressure, a pressure regulator to control the pressure of the air inlet, a metering valve to control the flow rate of the foam solution, and an adjustable damper to control the flow rate of the air. Foam solution flow rate through the nozzle is a function of pressure, therefore to ensure accurate flow rate data, the pressure readings should be taken as close to the nozzle as possible. Additionally, there is a solenoid valve for remote control, a bleed valve to depressurize the system during shutdown, and a liquid pressure gauge to take accurate pressure readings just before the nozzle. The transparency of the air cylinder is not a requirement, but could allow the nozzle and screen during operation to be visualized for better control. Many improvements have been performed based on the original model illustrated in NFPA11 for research purpose, such as the orientation of the transparent air cylinder, the inclusion of a pump, and the differences in the piping setup. A summary of major improvements is listed in Table 1. The new setup is shown in Fig. 2. The horizontal orientation of the transparent air cylinder performs two functions. The first is that a horizontal setup allows foam to be carried by the air flow toward a target. This adjustment allows the flexibility of developing foam that is deposited into a foam collection tank directly beneath the foam solution or projected toward a location 3e5 m away from the source. In order to direct the flow of foam, a reflector plate was installed at the end of the transparent air cylinder. The position of this plate can be adjusted to push the foam down at a variety of angles allowing the foam to be directed onto spills at various distances from the generator. The other function this orientation provides is the inclusion of the drip
175
tray. In the vertical setup, any foam solution that is not converted into foam will drip into the foam collection tank. In the new design, the excess foam solution settles in the drip tray which is positioned just after the screen. From there the foam solution moves through a hose back to the foam solution tank. The drip tray not only limits foam solution waste, but also helps in accurate characterization of expansion ratio and the effect of foam application on the vaporization rate of an LNG spill. Because the expansion ratio calculation is done by dividing the foam volume by the mass of foam solution used, any foam solution not converted to foam will increase the mass without increasing the foam volume, giving inaccurate data. When applied to cryogenic liquids, unconverted foam solution can also have detrimental effects, similar to the water discharge phenomenon. If unconverted foam solution is applied to an LNG pool, the large temperature difference between the foam solution and the cryogenic liquid pool and the high heat capacity of foam solution will cause additional LNG to evaporate, which is counterproductive to the application of foam and should be avoided in the study. The inclusion of a pump in the system is both a necessity and an improvement. During pressure testing, the foam solution tank, a 50 L LDPE carboy from US Plastic Corp, was found to withstand around 28 psig. However, the system pressure must be able to reach 50 psig to achieve the desired foam solution flow rate. To resolve this problem, the air supply was removed and a pump was added downstream of the foam solution tank to pressurize the foam solution. This new design also has some advantages over the previous design in NFPA 11. The first is that the new system only requires power to operate, meaning that it can be transported and used in areas that do not have a pressurized air source. The second improvement is that the pump can boost the pressure much higher than the pressurized air, allowing a higher maximum foam solution flow rate. Because a majority of the setup is made from stainless steel and the PVC components can withstand 160 psi, running at higher pressures is a viable option. The third improvement is that the new design is inherently safer and easier to operate. Because the pump is placed downstream of the foam solution tank, the pressurized section of the setup is smaller due to the exclusion of the tank. This modification makes the setup inherently safer due to a smaller pressurized volume. Furthermore, the removal of the pressurized air simplifies the startup and shutdown procedures, making operation of the equipment easier on the researcher. These two changes illustrate two of the principles of inherently safer design, minimize and simplify. The last improvement is that the drain can feed directly back into the unpressurized foam solution tank to minimize the foam solution losses, while in the original design, the drain would have to go to waste because it could not feed back into the high pressure system. The last major change is the difference in piping setup. In the original NFPA diagram, there is very little detail on how to actually assemble the apparatus. Many of the piping decisions were made
Table 1 Improvements of foam generator design. Original design in NFPA 11
New design in this work
Improvements
Vertical air cylinder
Horizontal air cylinder
Solution propelled by pressurized air Bleed valve used to release pressure during shutdown Solenoid valve used for remote startup and shutdown Vertical flow
Solution propelled by jet pump
Allows higher foam outlet point due to ceiling height restriction Improves safety of design and removes dependence on pressurized air Improves design safety and simplifies shutdown procedure With smaller pressurized volume, valve is no longer needed, startup and shutdown simplified Allows foam to be directed into foam collection tank for an even fill
Smaller pressurized volume (50 L in the original design to 1.5 L in the improved design), pressure naturally bleeds out the nozzle Removed Solenoid valve Inclusion of deflector plate or a plastic duct
176
B. Harding et al. / Journal of Loss Prevention in the Process Industries 39 (2016) 173e180
Closeup of Damper Damper Handle 100 % Open 50 % Open Unistrut® Frame (8' x 3')
10 % Open
Transparent Air Cylinder
Deflector Plate Nozzle
Airflow In
Elbow
Air Out
Nipple Screen Fan Adjustable Damper Foam SoluƟon Tank
Pressure Gauge
Foam Out
½” PVC Tubing
Ball Valve
Closeup of Screen Assembly Gasket Washer Washer Gasket
Tap
½” Stainless Tubing
¾” Stainless Tubing
Bolt Screen Nut Flange Flange
Pump Pressure Regulator
Fig. 2. Diagram of final high expansion foam generator.
during the planning and construction of the improved apparatus, some of which add to the functionality of the setup, and others which are simply required for the foam solution to flow through the system. One such addition is the 10 foot flexible PTFE/stainless steel hose. This addition allows the height of the air cylinder to be changed, permitting the application of foam into containers of different height or the application of foam at different trajectories. Another addition is the elbow/nipple combination located before the nozzle. This section of the piping allows nozzle repositioning in reference to the screen, ensuring that the screen is properly coated with foam solution and no solution is sprayed on the sides of the air cylinder. There are also multiple nozzles and multiple screens with different mesh sizes that can be selected depending on the application. The transparency of the air cylinder is not a requirement, but could allow the nozzle and screen to be visible during operation for better control. In order for the apparatus to be effective, it must be able to continuously apply high quality foam at the desired expansion ratio. In order to achieve this goal, many of the aspects of the apparatus must be adjustable to reach the desired foam application. The aspects of the foam that can be controlled are the foam flow rate, foam expansion ratio, and foam bubble size. Each of these aspects is controlled by adjusting multiple different components of the system. The foam flow rate can be changed by changing the nozzle
from a low flow nozzle to high flow nozzle and by increasing the pressure by turning up the pump and the pressure regulator. The foam expansion ratio can be changed by changing the foam solution flow rate as mentioned before in combination with the air flow rate from the fan. The foam bubble size can be changed by adjusting the mesh size of the screen and the air flow rate. 3. Characterization and discussion In order to test the validity of the foam generating apparatus, foam was produced and characterized. The three important characteristics of the foam are expansion ratio, time to half-height, and foam application rate. Expansion ratio is defined as the ratio of the amount of air to the amount of liquid foam solution in a foam. In firefighting, foams with an expansion ratio less than 20:1 are called low-expansion foams, between 20:1 and 200:1 are called mediumexpansion foams, and greater than 200:1 are called high expansion foams (NFPA, 2012). Each type of foam has advantages and disadvantages depending on the application scenarios. Low expansion foams can behave like low-viscosity liquids, which make them ideal for covering large areas quickly (NFPA, 2012). Also, due to the lower air content, and associated higher specific gravity, they are affected less by the wind and other adverse weather conditions (NFPA, 2012). Additionally, they can be blasted or thrown similar to water
B. Harding et al. / Journal of Loss Prevention in the Process Industries 39 (2016) 173e180
coming out of a fire hose (Martin, 2012). Because their density is much lower than that of water, they have much less inertia and therefore cannot be blasted as far, but they can still be used on fires that are far away. On the other end of the spectrum, high expansion foams are more useful when trying to fill a space, because of the high expansion ratio (Martin, 2012). If, for instance, there is a fire in a storage facility or a hangar, the foam can be used to quickly fill the structure without using large quantities of foam solution (Martin, 2012). High expansion foams are also ideal for use on LNG or other cryogenic liquid spills. The low liquid content of the foam contributes less latent heat to the spill while still providing the heat insulation and physical barrier effects. High expansion foams cannot be blasted because of their very low density. These foams must be produced and deposited directly on the area of application. Medium expansion foams have some of the applicability of each of the previous types. They can be blasted about 12e35 m, depending on the expansion ratio, making them very versatile (Sthamer, 2012). 3.1. Expansion ratio As mentioned before, expansion ratio is defined as the ratio of the amount of air to the amount of liquid foam solution in a foam. The two contributing factors to this ratio are the air flow rate and the liquid flow rate. The air flow rate is controlled by the damper, positioned behind the fan. In this study, the damper conditions of 100% open, 50% open, 25% open, and 12.5% open, as well as liquid flow rates between 1.4 and 5.7 L/min, were used. The liquid flow rate is controlled by the pressure regulator positioned directly after the pump. The foam collection tank is positioned on a balance, which is used to measure the mass change in the tank. With a constant liquid flow rate, increasing the air flow rate will cause the expansion ratio to increase, assuming the air is being trapped within the bubbles. With the mass change and the total volume of the foam collection tank, the expansion ratio can be determined. There is a very slight mass contribution from the air trapped in the foam bubble due to the increased pressure in the bubble, however this contribution is approximately 2.8 g, which represents a minimal effect and can be neglected. From Fig. 3, it can be found that all the foams have expansion ratio higher than 200, indicating they all are high expansion foams, and the foam expansion ratio decreases with increasing the liquid flow rate (affected by applied pressure) when maintaining constant
177
aperture size. Although only high expansion foams were created in these tests, using different conditions and a different foam solution could yield foams with much lower expansion ratios, if a lower expansion ratio was desired. An interesting result was the similarity of the 100% open and 25% open results. A possible explanation is that with the damper 100% open, the air flow is in excess of what it takes to create foam, and for this reason the foam being created in both tests is very similar. When running the test with aperture 100% open, it was observed that a considerable amount of air is not entrained in the bubbles, instead, it is used to carry the foam giving it a much straighter and more forceful horizontal trajectory. However, in the 25% open test, the foam follows a parabolic trajectory and easily flows downward into the foam collection tank. This difference in trajectory is due to the difference in air flow. Another interesting phenomenon is the higher expansion ratio of the 50% open damper position at low foam solution flow rate. Where 25% open and 100% open begin to flatten out, 50% continues its upward trend. This is probably due to multiple factors affecting the bubble formation. In the 50% open trials at 2.3 L/min and 3 L/ min, although the expansion ratio was higher than any other trials with the large nozzle, the foam stability was much lower than trials with a similar foam solution flow rate, indicating that something about the foam bubbles may have been fundamentally different, such as the thickness of the bubble wall. The bulk of the trials were run using the WL 1½ nozzle from BETE, which produces a flow rate of 5.7 L/min at 40 psi (BETE, 2014). In order to test even lower flow rates, which were expected to have the highest expansion ratios, a smaller nozzle, called the WL 1, was used. This nozzle produces a flow rate of 3.8 L/min at 40 psi. Trials to the left of the vertical line in Fig. 3 are using this smaller nozzle. Through the use of the smaller nozzle, an expansion ratio of 852 was achieved, which is among the highest in any of the trials. As the pressure drops below about 5 psi, the conical spray pattern of the nozzle narrows into a jet and does a poor job of covering the screen. Switching to a smaller nozzle allows the researcher to achieve a lower foam solution flow rate while maintaining a pump pressure above 5 psi. Trials were also conducted using a pressure of 11.5 psi with the WL 1 nozzle, which gives an equivalent foam solution flow rate to 5 psi using the WL 1½ nozzle. In all of these trials the expansion ratio was higher using the small nozzle than the equivalent flow rate using the larger nozzle, which can be seen in Fig. 4. The likely explanation is that by operating at a higher pressure, the spray pattern is able to cover a larger area of the screen and do so with more uniformity, allowing additional air to be entrained in the bubbles creating a higher expansion ratio foam. 3.2. Time to half-height
Fig. 3. Expansion Ratio vs. Foam Solution Flow Rate for all expansion foam trials. Values in the legend represent aperture position and are written as percentage of the area open to flow. Trials left of the vertical line were conducted using the smaller nozzle, all others were conducted using the larger nozzle.
The term “metric quarter time” is usually used to determine the rate of liquid drainage from the foam matrix (Conroy et al., 2013). Quarter time is defined as the time it takes one quarter of the foam solution to drain out of the foam matrix. In the current setup, draining rate cannot be measured, however the height of the foam in the container can be measured accurately. For this reason, the stability metric will be time to half-height, which is defined as the time the foam takes to settle to half of its original height. The foam addition and subsequent collapse has been captured by video camera for each experiment and analyzed to determine time to half-height. Each trial was filmed in order to collect real time height information for the collapse rate. The foam collection tank was marked with horizontal lines every six inches to provide a scale on the camera. The height at the center was used as the reference point (18 inches from each wall). Time to half-height is used as the main
178
B. Harding et al. / Journal of Loss Prevention in the Process Industries 39 (2016) 173e180
Fig. 4. Comparison of expansion ratio at different damper positions using the Small Nozzle (WL1) and Big Nozzle (WL1½) keeping foam solution flow rate constant.
comparison tool, however for certain trials with interesting characteristics, MATLAB was used to obtain height information on 1-s intervals and observe the entire foam height curve. For foam stability, assuming similar bubble size, foams with higher expansion ratios should have a shorter time to half-height, because there is less foam solution present in the initial foam. In this data the trend is present, but the correlation is very weak. In Fig. 5, two data points are highlighted, Trial A and Trial B. Both are run at identical foam generator conditions (Pump pressure 30 psi, damper position 25% open) and have similar expansion ratios (Trial A was 405, Trial B was 393), however their stability characteristics are dramatically different. The Trial A time to half-height was 282 min and the Trial B time to half-height was 151 min, representing a decrease of more than 46%. In Trial A, the foam solution had been mixed approximately 28 h before application, whereas in Trial B, the foam solution had been mixed approximately 220 h before application, giving the foam solution in Trial B much more time to suffer stability losses. The potential causes for this drop in stability are the age of the foam solution, the interaction of the foam solution with the pipe work, and the interaction of the foam
Fig. 5. Time to Half-Height vs. Expansion Ratio for all large nozzle trials. Trials A and B have been highlighted to show the effect of foam solution age on time to half-height.
Fig. 6. Time to Half-Height vs. Foam Solution Age for exposed and unexposed scenarios.
solution with air. These effects on foam stability prompted a series of experiments on measuring foam stability with foam solution age as the independent variable. The results of these experiments are shown in Fig. 6. In the first study, which focuses on the effect of foam solution age, the foam solution was kept in a sealed carboy to limit the amount of exposure to the air. These experiments show no decrease in time to half-height with increased age, meaning that the drop in foam stability was likely due to interaction with either air or the pipework. The second study was conducted with the foam solution left in the foam solution tank, which is not isolated from the air, and a significant drop in time to half-height occurred with foam solution older than eight days. This study also served to eliminate the potential factor of foam solution interaction with the pipe work due to the offset time in between trials. The first trial (at one day) was run immediately, with no time spent in the pipework and no time spent exposed to air. The second trial (at eight days) was run with seven days spent exposed to the pipework and seven days spent exposed to the air. The third trial (at 10 days) was run with two days spent exposed to the pipework and nine days spent exposed to the air. If pipework exposure was the determining factor in foam stability, then the time to half-height for the trial at 10 days
Fig. 7. Foam breaking info for a full trial with a linear trend line for visual reference.
B. Harding et al. / Journal of Loss Prevention in the Process Industries 39 (2016) 173e180
179
4. Conclusions 4.1. Main features and the benefits on foam related research
Fig. 8. Foam Production Rate (m3/min) for Expansion Foam Tests. Only trials using the larger nozzle are depicted. Values in the legend represent aperture position and are written as percentage of the area open to flow.
would be expected to be higher than the time to half-height at eight days. From these trials, the conclusion can be made that exposure to air is the main contributor to a decrease in foam solution stability. Fig. 7 shows the foam height throughout a single trial to illustrate how foam height changes with time. In this trial the pressure was 5 psi and the damper was set at 25% open, resulting in an expansion ratio of 643 and a time to half-height of 244 min. For the purposes of comparison, the linear fit predicts a time to half-height of 245.4 min. For this analysis, the video was broken into pictures at 1 min intervals and MATLAB was used to analyze the position of the upper surface of the foam. From that information, foam height could be calculated accurately across the trial. As the figure demonstrates, there is a highly linear dependence of height on time, with an R2 value of 0.978. This linearity is also present in other trials regardless of application conditions.
3.3. Foam application rate In addition to foam quality characterization such as expansion ratio and time to half-height, another important criterion of foam generation is foam application rate. The application time is the period between the moment the first bubble of foam entered the foam collection tank to the moment the foam generator was unpowered, recorded by a video camera. The foam application rate is determined by dividing the total volume of the foam collection tank by application time, the error of which is estimated to be within 4% due to a trivial part of foam spill out of the tank. Foam application rates for all of the trials were found to be between 1.2 and 2.2 m3/min, which are well suited for lab-scale research purpose, compared with the high application rates of commercial generators (larger than 38 m3/min). As shown in Fig. 8, the foam production rate for the damper position at 12.5% open was much lower than the production rate at other damper settings because the air flow rate was the limiting factor in foam formation. The foam production rate is essentially a measure of how much air is trapped in foam bubbles and when the air flow rate drops below a threshold value, the foam production rate drops along with it. This same effect can be seen in the expansion ratio data in Fig. 3, where the limiting air flow rate caused a decrease in expansion ratio when compared to other damper settings.
The foam generating apparatus discussed in this paper was able to address all of the issues that exist with using an industrial scale foam generator for lab use. The foam generator was able to produce high quality foams at flow rates that fit lab scale research. These foams had expansion ratios between 298 and 892, with an average time to half-height around 185 min. The uncoupling of the air flow rate and the liquid flow rate also allows the researcher customizability without changing components of the generator, which is an advantage over industrial scale foam generators. The portability and availability of the equipment, lack of water discharge at the beginning of a trial, and lower cost of building equipment are also advantages to the researcher that this equipment can offer. The constructed generator also exhibited good customizability, allowing the researcher to substitute many of the parts in order to create a foam with a variety of different physical characteristics. The main variables that were adjusted were foam solution pressure, damper position, and nozzle size, each of which showed a strong effect on the resulting foam. In addition to these variables, many other adjustments can be made to produce a desirable, applicable foam. The main improvements as they pertain to LNG research are the control over independent variables and the lack of water discharge at startup. With these improvements over the commercial foam generator, researchers will be better able to conduct trials with specific variables in mind and with a high control over the conditions present in the experiment, allowing a more in depth study of the factors related with foam application to LNG spills or other foam related research. References Angus Fire, 2014a. Fixed Turbex Systems (WWW Document). http://www.angusfire. co.uk/products/foam_equipment/fixed/turbexlng.html (accessed 6.13.14). Angus Fire, 2014b. LNG Turbex Skids (WWW Document). http://www.angusfire.co. uk/products/foam_equipment/fixed/lngturbexskids.html (accessed 6.13.14). Angus Fire, 2014c. Mini-turbex (WWW Document). http://www.angusfire.co.uk/ products/foam_equipment/portable/miniturbex.html (accessed 6.13.14). BETE, 2014. Open Performance Data Table - SI/Metric Units (WWW Document). http://www.bete.com/pdfs/BETE_WL-metric.pdf (accessed 2.8.14). Chemguard, 2014a. General Foam Information (WWW Document). http://www. chemguard.com/about-us/documents-library/foam-info/general.htm (accessed 4.2.14). Chemguard, 2014b. Water Powered High Expansion Foam Generator (WWW Document). http://www.chemguard.com/fire-suppression/catalog/EngineeredSystems/high-expansion-foam-generators/1500wp.aspx (accessed 6.13.14). Conroy, M.W., Taylor, J.C., Farley, J.P., Fleming, J.W., Ananth, R., 2013. Liquid drainage from high-expansion (HiEx) aqueous foams during and after filling of a container. Colloids Surf. A Physicochem. Eng. Asp. 426, 70e97. http://dx.doi.org/ 10.1016/j.colsurfa.2013.02.050. Cronce, D.T., 2002. Chemical Warfare Agent Decontamination Foaming Composition and Method, 6376436 B1. Hiltz, R., 1993. Foam Blanketing: The Use of Foam to Mitigate the Vapor Hazard of Spilled Volatile Chemicals. In: Prevention and Control of Accidental Releases of Hazardous Gases, pp. 216e231. Jamison, W.B., 1966. Fire-fighting Foam Generator, 3241617. Jamison, W.B., Barnes, R.W., 1965. Fire-fighting Method Employing High Expansion Foam, 3186490. Love, A.H., Bailey, C.G., Hanna, M.L., Hok, S., Vu, A.K., Reutter, D.J., Raber, E., 2011. Efficacy of liquid and foam decontamination technologies for chemical warfare agents on indoor surfaces. J. Hazard. Mater 196, 115e122. http://dx.doi.org/ 10.1016/j.jhazmat.2011.09.005. Martin, T.J., 2012. In: Stevenson, P. (Ed.), Fire-fighting Foam Technology. John Wiles & Sons, Ltd, Chichester, UK, pp. 411e457. NFPA 11: Standard for Low-, Medium-, and High-Expansion Foam, 2012. NFPA 24: Standard for the Installation of Private Fire Service Mains and Their Appurtenances, 2013. O'Regan, J.F., Lundberg, B.A., Mussoni, W.J., 1970. High Expansion Foam Generator, 3512761. Sthamer, R., 2012. Medium Expansion Foam (WWW Document). http://sthamer. com/englisch/f32_medium_foam.html# (accessed 10.21.13).
180
B. Harding et al. / Journal of Loss Prevention in the Process Industries 39 (2016) 173e180
Tucker, M.D., 2008. Enhanced Formulations for Neutralization of Chemical Biological and Industrial Toxants, 7390432 B2. Tucker, M.D., Betty, R.G., Tadros, M.E., 2004. Concentrated Formulations and Methods for Neutralizing Chemical and Biological Toxants, 6723890 B2. Williams, E.C., 1953. Apparatus for Extinguishing Fires, 2645292. Zhang, B., Harding, B., Liu, Y., Mannan, M.S., 2015. Mitigation effect of high
expansion foam on LNG vapor hazard. In: 11th Global Congress on Process Safety. Austin, Texas. Zhang, B., Liu, Y., Olewski, T., Vechot, L., Mannan, M.S., 2014. Blanketing effect of expansion foam on Liquefied Natural Gas (LNG) spillage pool. J. Hazard. Mater 280, 380e388. http://dx.doi.org/10.1016/j.jhazmat.2014.07.078.