18.10.11 I am only updating the published papers section of this web-page. I don't have time to update the content!
The phenomena of spontaneous ignition due to internal heating in bulk solids such as coal, grain, hay, wool wastes etc. can be described by thermal explosion theory as developed by Semenov and Frank-Kamenetskii (Frank-Kamenetskii, 1969; Bowes, 1984). In such models heat release is usually represented by a single Arrhenius reaction and combustion is initiated when heat-loss is unable to balance heat generation by the internal heating of the bulk material. However in industrial processes involving large volumes of bulk organic materials there are two sources of heat-generation: a low-temperature process involving the growth and respiration of micro-organisms, such as aerobic mould-fungi and bacteria, and a high-temperature process due to oxidation of cellulosic materials. We are investigating simple models to predict the thermal behaviour of cellulosic materials in the presence of micro-organisms (biomass) undergoing exothermic reactions.
Examples of processes where biological heating is important include the use of large scale composting operations as a significant biorecycle process (Rynk, 2000), the storage of industrial waste fuel, such as municipal solid waste (Hogland et al, 1996), and landfill sites. In these systems, self-heating due to biological activity is an inherent consequence of the process and normally a goal, e.g. in composting (Brinton et al, 1995). Elevated temperature of the order 70-90C may be found within a few months or even a few days (Hogland et al (1996)). Municipal solid waste may not seem an obvious source of combustible materials, however in one set of experiments approximately 85% of industrial waste was deemed to be combustible (Hogland et al, 1996).
Although it has been recognised for over twenty years that
"biological heating may be an indispensable prelude to
(Bowes, 1984:page 373),
very little information is available regarding the mechanism of fires
when biological self-heating is involved. An understanding of this
phenomenon is crucial as fires (most likely due to biological
self-heating) are common at landfills worldwide
Furthermore, spontaneous combustion may be the most frequent cause of
fires at compost facilities
In additional to the industrial process mentioned previously it has long been suspected that the initiation of haystack fires is due to biological heating under the influence of water content or air humidity, which then lifts the temperature to be sufficiently high to trigger significant oxidation of cellulose material and consequently leading to fire. So far, however, only the influence of oxidative processes has been considered.
We model the heat release rate due to biological activity as a function which exhibits two types of behaviour: over the temperature range 0<= T<= a it is a monotonic increasing function of temperature whereas for T>= a it is a monotone decreasing function of temperature. This function is given by
(1) k(T) = A1exp[-E1/(R*T)]/ (1+A2exp[-E2/(R*T)].
This formula encapsulates that activation and inactivation processes occur over different temperature ranges. At low temperatures the metabolic activity of the biomass increases with increasing temperature as enzyme activity raises. These processes are governed by the growth parameters A1 and E1. However, for sufficiently high temperatures the essential cell proteins which are heat sensitive start to denature leading to cell death. These processes are represented by the biomass deactivation parameters A2 and E2. To ensure that the heat release due to biological activity has a global maximum with respect to temperature, the activation energy for the inhibition process must be larger than the activation energy for the biomass growth, i.e. E2 > E1.
can be derived theoretically by assuming that the biomass
growth rate is determined by a rate limiting step in which there is an
equilibrium between activated and unactivated forms of the biomass
(Roels, 1983, Chapters 7.5.2 and 9.5).
From this, thermodynamic, perspective the term
A2exp[-E2/RT] represents the temperature
dependence of the equilibrium constant whilst the expression
A1exp[-E1/RT] is the maximum forward
rate of reaction in the rate limiting step.
Roels (1983, page 253) notes that
``Although this model is based on a highly simplified image of the complexity
of the growth process, it can be considered an efficient tool to model
the temperature dependence of the maximum rate of growth''.
Equation (1) has been used to model the maximum specific biomass growth rate in the aerobic biodegradation of the organic fraction of municipal waste (Liwarska-Bizukojc et al, 2001). It has also been used in a number of models for solid-state fermentation processes (Khanahmadi et al, 2004; Mitchell and von Meien, 2000; von Meien and Mitchell, 2002; von Meien et al, 2004).
In practice there is not a unique microbe reasonable for heat generation in a compost pile, but rather many different species which thrive over a sequence of overlapping temperature intervals (Kubler, 1987). The temperature varying active biomass concentration temperature is implicitly incorporated into the model through the chosen functionality for the biological heat release rate.
24.08.11. I have stopped updating this part of the web-page. Too many other things to do!
Analysed a model for self-heating in compost piles (Nelson et al 2003), based upon Semenov's theory for thermal explosions. This showed that the interaction between energy released from the biological and chemical reaction can lead to one of three types of behaviour: negligible temperature increase; an elevated temperature increase without ignition; and ignition of cellulosic materials. The elevated, but non-ignition, temperature branch is the feature of practical interest in facilities such as industrial compost facilities and municipal tips. Mathematically, this branch is generated by a Quartic Fold. The three solution branches is reminiscent of smouldering combustion.
Another interpretation of our results is that it is possible for the heat released by biological activity to raise the local temperature sufficiently high that ignition of the cellulosic material occurs. Thus a situation which is judged subcritical, i.e. safe, based on classical thermal explosion theory may be supercritical. This has implications for safety!
Since the system investigated was described by a single (but non-linear), first-order ordinary differential equation, with few parameters, it is possible to thoroughly investigate the generic steady-state behaviour of the system when parameters are varied using singularity theory.
This paper was awarded a `John A Brodie Medal: Certificate of Merit' for being one of the top six papers published in the 9th Asia Pacific Confederation of Chemical Engineers Congress (2002) in practical application of chemical engineering. They were over 900 papers presented at the congress.
The Frank-Kamenetskii variables were used and the pre-exponential approximation was made. The model equations then only contain three parameters and it is possible to thoroughly investigate the generic behaviour of the model. For both 1-d slab and 2-d rectangular slab geometries we showed that there are two generic steady-state diagrams, including one in which the temperature-response curve is the standard S-shape curve familiar from combustion problems. Thus biological self-heating can cause elevated temperature raises due to jumps in the steady temperature.
This was problem was used to test a recently developed semi-analytical technique. For the 2-d problem a four-term expansion is found to give highly accurate results - the error of the semi-analytical solution is much smaller than the error due to uncertainty in parameter values. We conclude that the semi-analytical technique is a very promising method for the investigation of bifurcations in spatially distributed systems.
Made a preliminary numerical investigation into a spatially dependent model containing biological and chemical activity based upon thermal explosion theory (Sidhu et al 2006, 2007a). The assumptions of the model include the following.
Steady-state analysis shows that there are three stable branches and two unstable branches. Two of the stable branches correspond to `low temperature' solutions. The lowest of these represents a state of negligible heat-reaction which is undesirable in applications. The higher of the `low-temperature' branches is an elevated temperature branch, which is the feature of interest in industrial composting facilities and municipal tips. There is also a much higher temperature branch which represents flaming combustion within the heap. Transitions between the stable solution branches are defined by limit-points. The system moves from the negligible heat-reaction branch to the elevated temperature branch as the width of the compost pile is increased through a value corresponding to a low temperature limit point (LLP). A corresponding transition from the elevated temperature branch to the negligible reaction branch occurs as the width is decreased through an extinction limit point (ELP). The transition from the elevated temperature branch to the flaming combustion branch is associated with a high temperature ignition limit point (HLP). (There is also an extinction limit point on the flaming combustion branch, but this has little practical significance).
The region in which the (desirable) elevated solution branch is possible narrows as the ratio of the height to total length of the compost pile increases. At a critical value of the ratio there is a double limit-point bifurcation (at which the width at which the ELP and HLP occur are identical). For values of the ratio greater than the critical value increasing the width of the compost pile through the value at the LLP sees the temperature profile in the compost heap evolving not to the elevated temperature branch but straight to the flaming combustion branch. Similar behaviour was observed in the spatially uniform Semenov model analysed by Nelson et al (2003).
Our results suggest that composting performance can be improved by insulating the lower boundary.
Made a preliminary numerical investigation into a spatially dependent model containing biological and chemical activity that extended the model based upon thermal explosion theory (#2 above) to include consumption of oxygen (Sidhu et al 2007b). The model assumptions are essentially as stated for the thermal explosion model (#2 above). We compared the bifurcation behaviour in models that included and excluded oxygen in both one- and two-dimensional geometries.
The following modifications were made to the model.
We found that the models without/with oxygen consumption predicted very similar values for the low temperature limit point (LLP) and the extinction limit point (ELP). The model predictions were in good agreement upto the critical value of the compost length that defines the high temperature ignition limit point (HLP) in the model without oxygen. Recall from (#2) above that in the model without oxygen the HLP represents a transition point at which flaming combustion occurs within the pile. The HLP was not found in simulations of the oxidation model. Above the length corresponding to the HLP bifurcation in the thermal explosion model the temperature of the oxidation model increased very slowly. These scenario could change if there is a steady supply of oxygen through the pile.
|Dr. E. Balakrishnan.||2000-2007|
|Professor X. Dong Chen||2000-Present|
|T. Luangwilai (PhD student)||2008-Present|
|Dr H.S. Sidhu||2005-Present|
|Professor G.C. Wake||2000-2007|
This paper was awarded a `John A Brodie Medal: Certificate of Merit' for being one of the top six papers published in the 9th Asia Pacific Confederation of Chemical Engineers Congress (2002) in practical application of chemical engineering. They were over 900 papers provided to the congress. The award is given annually by the Chairman of College of Chemical Engineers, The Institution of Engineers of Australia.
The propensity of coal to undergo self-heating and spontaneous combustion is a major problem wherever coal is being mined, stored or transported, posing problems for both coal producers and users. Accordingly it has been the subject of extensive fundamental and practical research for well over a hundred years.
A succinct overview of the issues relating to the spontaneous combustion of coal and coal mine fires was provided by Banerjee (1985). Carras and Young emphasised the causes of self-heating, described mathematical models for modelling self-heating and assessed the limitations of industrial test methods. Recently Babrauskas [pages 719--724] has provided an overview of the ignition of coal, with a particular interesting description of the historical development of the subject towards the end of the 19th century. Our understanding of the chemical reactions that occur during the low-temperature oxidation of coal was reviewed by Wang et al (2003).
Experimental work covering the period 1996-2005 was reviewed by Nelson & Chen. The introduction of this article appears in the following paragraphs.
In this article we review experimental work covering the period 1996-2005. We start in section 1.1 by outlining the economic and environmental consequences of unwanted coal fires and the threat that these fires pose to human life. In section 1.2 we describe the phenomenon that is frequently the reason for coal fires: spontaneous combustion. The connection between spontaneous combustion and coal fires has motivated the ongoing stream of publications relating to the spontaneous combustion of coals.
Although the emphasis of this paper is experimental work, it will be useful to have an understanding of some of the issues involved in modelling the spontaneous combustion of coal. These are covered in section 2. In particular, in section 2.2 we discuss the simplest theory for the spontaneous combustion of bulk materials subject to self-heating; namely the model developed by Frank-Kamenetskii (1969). This allows some elementary insights into the factors governing spontaneous combustion and provides the basis for an experimental method, the hot-storage test, discussed in section 16, to determine kinetic parameters. These parameters are required as inputs for any mathematical model of the self-heating process. In section 3 we review some issues dealing with the kinetics of the oxidation of coal at low temperatures.
In sections 4-9 we discuss the spontaneous combustion of coal in various environments. In sections 4 and 5 we consider spontaneous combustion in the context of `natural' coal fires and coal mine fires respectively. By a `natural' coal fire we mean a fire in a coal seam that is not being worked in a coal mine. The distinction between these two categories is very fine, but in view of the extra issues associated with coal mine fires, such as means of detecting self-heatings at an early stage and suppression of combustion, we think that it is useful. Fires in abandoned coal mines are discussed in section 4.1. In section 6 we consider the spontaneous combustion hazard posed by accumulating layers of coal dust. In sections 7 & 8 we discuss hazards associated with stockpiles of coal. Section 7 deals with stockpiles of `fresh' coal, such as are found at power generators. Section 8 deals with stockpiles of the coal waste produced in mining operations. In section 9 we discuss the risk of spontaneous combustion when coal is transported.
In section 10 we discuss various properties of coal that influence its propensity to spontaneous combustion. The inherent moisture content of a coal and the humidity of the air are two properties that exert a strong influence over self-heating in a coal. There are discussed in section 11. Low-rank coals contain large amounts of water, which must be removed to make their use economic. However, as discussed in section 11, very dry coals have a high propensity to spontaneous combustion. Thus low rank coals must be treated to both decrease their water content and their propensity to self-heating. This is the topic of section 12.
In sections 13-18 we consider experimental techniques for investigating propensity to spontaneous combustion. We start in section 13 by outlining the various experimental procedures that exist for this task. In sections 14-18 we consider some of the more important methods: adiabatic methods, in section 14, the crossing point temperature, in section 15, the hot storage test, in section 16, the heat release rate method, in section 17, and the transient method, in section 18.
We draw our review to an end in section 19. In appendix 1 we define abbreviations used in this review and the nomenclature used in equations. In appendix 2 we provide a list of published parameter values for coals, including kinetic values. We hope that this collection of data will be useful in future modelling studies.