Analysis of Hay Clinker as an Indicator of Fire Cause Andrew T. Tinsley, MS, EI, CFEI Eastern Kentucky University, USA Michael Whaley, Criminal Investigator Tennessee Department of Agriculture, USA David J. Icove, Ph.D., PE, CFEI University of Tennessee - Knoxville, USA ABSTRACT The investigation of hay fires has long been a challenge for the fire service. Fires of this type are notoriously difficult to extinguish and usually require allowing the fire to run out of fuel (or the use of heavy equipment and large volumes of water). Inherently, this creates a black hole for fire investigators as they are often left with little more than witness statements to base their conclusions on. As a result, many hay fires are attributed to spontaneous combustion for lack of a better explanation. One of the traditional indicators of spontaneous combustion that fire investigators have relied upon in the past is the formation and/or presence of hay clinkers. Several reliable sources indicate the formation of hay clinkers is an event which is mutually exclusive to spontaneous combustion. After a string of suspicious cases in which hay clinkers were discovered, the Tennessee Department of Agriculture s Criminal Investigation Division conducted a series of field tests. The results of these field tests indicate that hay clinker production is possible with an external ignition source and should not be utilized as an indicator of fire cause. INTRODUCTION The investigation of hay fires has long been a challenge for the fire service. Fires of this type are notoriously difficult to extinguish and usually require allowing the fire to run out of fuel (or the use of heavy equipment and large volumes of water). Inherently, this creates a black hole for fire investigators as they are often left with little more than witness statements to base their conclusions on (see Figure 1). As a result, many hay fires are attributed to spontaneous combustion for lack of a better explanation. One of the traditional indicators of spontaneous combustion (or autoignition) that fire investigators have relied upon in the past is the formation of hay clinkers. Recently, there has been reason to call into question the reliability of the use of clinker formation as a valid indicator of fire cause or origin. Mike Whaley, Criminal Investigator for the State of Tennessee, Dept. of Agriculture, Agricultural Crime Unit, brought this issue to attention while investigating two barn fires in Coffee County, TN. Investigator Whaley was given reason to suspect arson. After discovery and identification of the clinkers, Investigator Whaley discovered several inconsistencies from conventional fire investigation texts and teachings. A series of tests were initiated to further investigate the formation of hay clinkers and their possible use as an indicator of fire cause.
Figure 1: Typical Barn Fire (Coffee County, TN) BACKGROUND Spontaneous Combustion Spontaneous combustion is defined as the general phenomenon of an unstable (usually oxidizable) material reacting with the evolution of heat, which to a considerable extent is retained inside the material itself by virtue of either poor thermal conductivity of the material or its container 1. In the case that this process leads to a flaming combustion it can be referred to as spontaneous ignition. A constituent component of the ability of a material to spontaneously combust is its ability to self heat. Babrauskas defines self heating as an increase in temperature due to exothermicity of internal reactions 2. The Fire Protection Handbook defines self heating (which it calls spontaneous heating) as the process whereby a material increases in temperature without drawing heat from its surroundings 3. There are many intricacies of spontaneous combustion and its ability to cause ignition of certain materials. Reviews of the process by which this takes place can be found in NFPA s Fire Protection Handbook 3, SFPE s Handbook of Fire Protection Engineering 1, and Babrauskas Ignition Handbook 2. There are several variables that will enhance a certain material s ability to spontaneously combust (and eventually ignite). These include 1 : The size of the body of material the larger the body of material, the more difficult it will be for the material to dissipate the heat generated by the selfheating process. High ambient temperature high ambient temperature is conducive to an acceleration of bacterial and fungal activity which often is a root cause of selfheating. A high ambient temperature also decreases the amount of energy lost to the surrounding atmosphere. Thermal Insulation the better a material s ability to retain heat, the sooner the temperatures in the material will rise above ambient. Fibrous Nature and Porosity of Material the porosity of the material (including the density of packing/stacking) affects the availability of oxygen, which is necessary for the biological reactions as well as the eventual combustion process. Temperature of Stacking temperature tends to accelerate the rate of biological reactions and decomposition.
Length of Time Undisturbed the longer the material is left undisturbed, the more time the self-heating reactions have to generate the necessary temperatures to produce a thermal runaway which can take on the order of months to occur. Gorbett and Pharr include a visual description of the above variables and their impact on the likelihood (or possibility) that spontaneous ignition will occur. It truly takes a specific set of circumstances and conditions for spontaneous ignition to occur. Figure 2 shows a graphical representation of the variables influencing the occurrence of a spontaneous ignition. As this paper primarily focuses on the spontaneous combustion of hay, it is necessary to understand the method of self heating which occurs in a porous, organic material. Babrauskas has an in depth literature review relating to cases of hay fires and studies of its ability to spontaneously ignite. According to the University of Tennessee s Agricultural Extension Service, hay fires will typically occur within 6 weeks of baling 4. These fires can occur in loosely or tightly packed bales as well as in hay stored indoors or outdoors. When hay is cut, it is not immediately dead. It takes a certain amount of time for the cells to die and the moisture content to drop to a level appropriate for baling and storage. If the hay is baled too early, the living cells continue to respirate and a small amount of heat is developed. If the bales are stored in such a fashion and density that the hay is unable to dissipate the excess heat, temperatures can rise to a point conducive to excess bacterial growth. Under normal conditions, the process would exhaust all excess moisture and the process would stop. However, if there is excess moisture in the hay is not consumed, bacterial and other microbial activity can continue eventually leading to an internal temperature of 54-60 C (130 to 140 F) at which the heat tends to kill the microorganisms 4. In the presence of thermophilic (heat loving) bacteria this process can continue to temperatures upwards of 71-77 C (160-170 F). At this temperature, the hay is converted to a form similar to a carbon sponge with microscopic pores 4. In this form and with the elevated temperatures present in the hay, combustion is now possible as long as there is enough oxygen present to support combustion (whether it is flaming or not). Figure 2: Conditions Required for Spontaneous Ignition to Occur in Materials Capable of Self Heating 5
Hay Clinker According to Dehaan, hay clinkers are a glassy, irregular mass, gray to green in color, composed of the inorganic residues of silicon, sodium, and calcium from the plant stems 6. Hicks performed an in depth review of hay clinkers, their constituent materials, and their formation. Hicks analysis tends to confirm the definition presented by Dehaan 7. Figures 3 and 4 show representative pictures of hay clinkers found at an actual barn fire in Coffee County, TN. According to Hicks, hay clinker is formed when silica (which can be accumulated in the hay via windblown dust, dirt, mud, etc. entrained during the baling process) melts and acts as a solvent to other refractory compounds in the hay 7. This melted silica and dissolved minerals will run to a low point and accumulate in a puddle. After drying, this puddle will turn into what we know as a hay clinker 7. Figure 3: Large Hay Clinker Figure 4: Average Hay Clinker
There is some debate as to whether the presence of hay clinker is an indication of the means by which the fire was started (i.e. the cause). Dehaan states The mechanisms thought to be responsible for hay clinkers are such that external ignition (such as an intentional fire) would not be capable of creating them. 6 Babrauskas states that These clinkers are not a suspicious device, but rather are formed from the minerals contained in the plants. 2 Hicks never truly states that the formation of hay clinker is an indicator either way, but does allude to hay clinkers as an indicator that spontaneous combustion may have occurred. 7 INVESTIGATION As shown in the literature review, there seems to be a clear difference in beliefs as far as the implications of the presence of clinker in a hay fire is concerned. Both DeHaan s and Babrauskas s texts are heavily relied upon in the fire investigation field at all levels of the profession. In order to investigate the development of clinker in conditions similar to arson, a series of tests were performed in an open field in Manchester, TN. Hay bales were purchased from a local farmer who had recently lost hay to a supposed arson fire. Several types of hay were used in the analysis. There was no analysis performed on the types of hay to determine its exact chemical makeup or type of grass. All grass tested was common in Middle and East Tennessee. Tests were performed in October 2008, September 2009, and November 2009. Table 1 contains a summary of the test dates, number of piles of hay burned, and the average ambient weather conditions over the duration of the test time. The tests were rather simple recreation including a stack of bales placed on a typical plastic pallet used for hay storage. The bales were covered with barn metal to simulate the collapse of the barn on top of the material after the fire had burned. Figure 6 shows a picture of the test prior to ignition of the fire. Figure 5 shows the fire approximately half way through the test. Eight individual piles were burned through the series of tests. The only variation in the tests was the inclusion of thermocouple monitoring into several of the piles. This data will be discussed later. Upon allowing the test to settle and cool enough to dig the scene, a variety of size and shapes of hay clinkers were developed. The results were indicative of the randomness of their creation as some piles produced clinkers while others did not. All clinkers found throughout the piles varied in size, color, and location. In addition to the tests described above, Jason Griner of the US Army performed an analysis of the material using FTIR (Fourier Transform Infrared Spectroscopy) Analysis as well as a standard microscope examination. After analysis, he concluded that the material is heterogeneous in nature composed of consistently inorganic materials very similar to minerals 9. According to Officer Griner, There is nothing that really stands out from a forensic perspective that could really tell you one hay clinker from another. 9
Figure 5: Test setup Figure 6: Test while burning Table 1: Test Conditions Date Number of Tests Ambient Temperature, C ( F) Relative Humidity (%) Wind Speed, km/h(mph) 10/31/08 1 65-68 28% 2.5-5.3 9/24/09 4 70-71 68% 2.0 11/03/09 1 60 41% 7.5 In order to fully investigate the development of clinkers within hay fires, the researchers felt it was necessary to monitor the interior temperatures of the fire itself. To accomplish this, a test setup shown in Figure 7 was implemented. The setup included four thermocouples held up by a metal fence post in the center of the pile of hay. The thermocouples were positioned at 25.4 cm (10 ), 53.3 cm (21 ), 81.3 cm (32 ), and 111.8 cm (44 ) from the bottom of the stack. Two separate piles were monitored this way, including one fire ignited externally (with a torch) and one fire ignited internally (to attempt to recreate the conditions seen in a spontaneous combustion fire). The internal fire was ignited by placing smoldering embers from a previous pile in the middle of the stack. The temperature distributions seen in the hay stacks are shown in figures 8 through 10.
Temperature, C Due to a thermocouple connection issue, the data for the middle bottom temperatures is unavailable. From the thermocouple data it is obvious that when the fire is started internally, the internal thermocouples (bottom and top middle) appear to develop higher temperatures faster. The internal fire did not appear to develop the maximum temperatures that the external fire developed. This is expected to an extent as the air entrainment for the internally ignited fires would be hindered when compared to the externally ignited fire. The maximum temperature seems to be much higher in the externally ignited fire as well. Figure 7: Thermocouple Setup Top Thermocouple 1800 1600 1400 1200 1000 800 600 400 200 0 0 200 400 600 800 1000 Time (s) External Internal Figure 8: Temperature Development 111.8 cm (44 ) from Ground Level
Temperature, C Temperature, C 1800 1600 1400 1200 1000 800 600 400 200 0 Middle Top Thermocouple 0 200 400 600 800 1000 Time (s) External Internal Figure 9: Temperature Development 81.3 cm (32 ) from Ground Level Bottom Thermocouple 1600 1400 1200 1000 800 600 400 200 0 0 200 400 600 800 1000 Time (s) External Internal Figure 10: Temperature Development 25.4 cm (10 ) from Ground Level DISCUSSION OF RESULTS It is known from the literature that hay fires started via spontaneous combustion can produce hay clinkers. The research provided in this paper has also shown that hay clinkers can be present in externally ignited hay fires. As indicated by Babrauskus, the presence or lack of hay clinker has nothing to do with the method by which the fire was ignited. In fact, the presence (or lack thereof) of hay clinker would appear to be related to the chemical makeup of the hay itself coupled with any impurities that may be introduced during the baling process. The location of hay clinkers within the fires were discovered throughout the test burns and actual barn fires. The clinker did tend to be concentrated near the centers of the hay stack, regardless of the source or location of ignition. It would also appear that the location of clinker within a fire would provide little insight into the point of origin in the fire. All in all, the presence of clinker appears to be an unreliable source of information regarding the cause and origin of a fire and should not have any weight placed on it.
RECOMMENDATIONS There are other special considerations that must be made in the investigation of hay fires. If available, moisture contents and internal temperatures of other hay put up around the same time can provide some insight to the cause of the fire, especially if they are stored under similar conditions. For untreated hay, moisture contents of 15-18% (for round bales) or 20-25% (for square bales) are viewed as the optimum moisture content prior to baling 4. The internal temperature should never exceed 60 C (140 F) 4. If the moisture content or temperature is above these limiting values, there is an increased likelihood that the fire would have been started by spontaneous combustion. Of course an investigation into the list of variables affecting spontaneous combustion can also provide insight into the cause of hay fires. The time since the hay was baled can also provide insight into the cause. The sources tend to vary on the time until ignition via spontaneous combustion, but they tend to focus on a range of from 3-10 weeks from the time it was baled. Under optimum conditions, hay fires can develop much quicker than the 3 week benchmark so this is not a foolproof method either. CONCLUSIONS Hay fires will continue to occur and will continue to be a problem for society as a whole. Investigations into these fires should be conducted using the same care and diligence that a structure fire would be investigated. With the disproof of hay clinkers as a viable indicator, other methods described in this paper and in the other fire investigation texts must be relied upon to reach an accurate conclusion for cause and origin analysis. As concluded in the discussion section of this paper, clinker should not be relied upon as a viable indicator of cause and origin. In fact, the analysis provided from Officer Griner appears to imply that it would be difficult to even distinguish hay clinkers developed in different fires. However, the traditional methods advocated in NFPA 921 still provide adequate means for guidance in the investigation of barn fires of all types. ABOUT THE AUTHORS Andrew Tinsley is an Assistant Professor with Eastern Kentucky University s Fire and Safety Engineering Technology Program as well as a Certified Fire and Explosion Investigator. He is a licensed Engineering Intern having received both his Bachelor s and Master s Degree from the University of Tennessee, Knoxville. He is currently pursuing his Ph.D. in Civil Engineering from the same institution. A resident of Richmond, KY, Mr. Tinsley has been active in the fire service for approximately seven years holding positions up to and including training officer with the Karns Volunteer Fire Department in Knoxville, TN. He can be contacted at andrew.tinsley@eku.edu. Mike Whaley is a Criminal Investigator with the State of Tennessee, Department of Agriculture, Agricultural Crime Unit. Inv Whaley specializes in wild fire investigation for the Ag Crime Unit and has been doing investigating of wildfires for 23 years. Inv Whaley is a certified arson investigator for the state through his in-depth fire training through various federal and state schools. Inv Whaley resides in Manchester, TN. and is married with two kids. Inv Whaley investigates wildfire arson, livestock theft and other farm crimes, and criminal activities on state forest lands for eleven counties in Tennessee. Inv Whaley just recently received the 2010 investigator of the year award.
David J. Icove, PhD, PE, CFEI, is a Research Professor in the Department of Electrical Engineering and Computer Science at the University of Tennessee, located in Knoxville, Tennessee. Dr. Icove is the co-author of several leading textbooks in the fire investigation field. He is a Registered Professional Engineer and a Certified Fire and Explosion Investigator (NAFI) and can be contacted at icove@utk.edu. ACKNOWLEDGEMENTS The authors wish to thank the State of Tennessee Department of Agriculture, Agricultural Crime Unit for providing the funding for the experiments. We also would like to thank Randy Baker for allowing us access to his equipment, property, and hay to perform the experiments. ENDNOTES 1. 2. 3. 4. 5. 6. 7. 8. 9. DiNenno, P.J. et al (eds) 2002, SFPE Handbook of Fire Protection Engineering 3 rd edition, National Fire Protection Association, Quincy, MA. Babrauskas, V. 2003, Ignition Handbook, Fire Science Publishers, Issaquh, WA. Cote, A. E. et al (eds) 2008, Fire Protection Handbook 19 th edition. National Fire Protection Assocation, Quincy, MA. Prather, T.G. 1988, Hay Fires: Prevention and Control, The University of Tennessee Agricultural Extension Service, Publication PB1306, October 1988. Gorbett, G.E. and Pharr, J.L. 2010, Fire Dynamics, Brady Fire, Pearson Prentice Hall, Upper Saddle River, NJ. DeHaan, J.D. 2007, Kirk s Fire Investigation 6 th edition, Pearson Prentice Hall, Upper Saddle River, NJ. Hicks, A.J. 1998, Hay Clinker as Evidence of Spontaneous Combustion, Fire Arson Investigator, July 1998, pp. 10-13. National Fire Protection Association 2008, NFPA 921: Guide for Fire and Explosion Investigations, National Fire Protection Association, Quincy, MA. Griner, Jason, Personal Communication, March 25, 2010.