Kömürün kendi kendine ısınmasının ve tutuşmasının modellenmesi, simülasyonu ve deneysel incelenmesi

Abstract

Kömürün ocak koşullarında kendi kendine ısınması, tutuşması ve yanmasının tahmin edilmesi henüz tam çözülmüş bir problem değildir. Olaya etkiyen çok sayıda parametre vardır. Özellikle kömürün kendi kendine tutuşmaya yatkınlığı/reaktivitesi ve içerisindeki mevcut orijinal nemin miktarı birinci derecede etkili parametrelerdir. Bu tez kapsamında ocak koşullarına mümkün olduğu kadar yakın koşullarda, kömürün kendi kendine ısınması ve tutuşması teorik ve deneysel olarak incelenmiştir. Kömür içindeki nemin etkisi dâhil pek çok parametrenin etkisini göz önüne alabilen yeni bir model geliştirilmiştir. Deneysel çalışmalar için adyabatik fırın deney düzeneği ve thermo – balance deney düzeneği imal edilmiştir. Geliştirilen kontrol algoritması ile adyabatikliğin daha kritik olduğu düşük sıcaklıklarda 0.1oC'ye kadar adyabatik ortam elde edilmiştir. İşletmelerin yardımıyla elde edilen ve kısa analizi yapılan numuneler üzerinde önce R70 ardından da kuluçka deneyleri yapılmıştır. R70 deneyi ile çok riskli grupta bulunduğu belirlenen A numunesinde, kuluçka deneyinde de 40 saat sonunda ısıl sürüklenme gözlemlenmiştir. Ancak riskli olduğu düşünülen B numunesinde özgün nem içeriğinin buharlaşma süreci sıcaklık artışını baskılamış ve ısıl sürüklenme gözlemlenmemiştir. Dolayısıyla kömürün içsel reaktivitesinin bir sonucu olan R70 deneylerinin tek başına yeterli olmadığı ve nem etkileşimlerinin derinlemesine incelenmesi gerektiği sonucuna ulaşılmıştır. Kömürün kendi kendine tutuşmasını, oksidasyon ve buharlaşma arasındaki ısı dengesi belirler. Bu nedenle modelde reaksiyon kinetiği ve buharlaşma/nem çıkışı kinetiği ayrıntılı olarak ele alınmıştır. Fiziksel modelde, ise adyabatik fırın koşulları esas alınmıştır. Bu çalışmada, yaşlanma etkisi dahil edilmiş yeni bir reaksiyon kinetik modeli önerilmektedir. Önerilen model için gerekli parametrelerin deneysel olarak nasıl elde edileceği 3. Bölümde ayrıntılarıyla verilmiştir. Kuruma – nem çıkışı kinetiği ifadesi ise temel kütle taşınım denklemi, monomoleküler seviyede bağlı nemi de içerecek şekilde düzenlenerek elde edilmiştir. Gerekli parametreler Reynolds sayısı sıfıra yaklaşırken bir küre etrafındaki akış için geçerli olan kütle geçişi bağıntılarıyla hesaplanmıştır. Tezin simülasyon bölümünde, öncelikle önerilen reaksiyon kinetiği modeli, R70 deneyleri kullanılarak doğrulanmıştır. Kuruma kinetiği modelinin doğrulanması için thermo – balance deneyi ile 80oC'de gerekli parametreler elde edilmiş, diğer sıcaklık ve debilerde modelle elde edilen sonuçlar deneysel sonuçlarla karşılaştırılmıştır. Teklif edilen teorik ifadeler, deneysel olarak ölçülen veriyi oldukça başarılı bir şekilde tahmin etmiştir. Geliştirilen simülasyon programı kömür numuneleri için yapılan kuluçka deneylerinin simülasyonu için çalıştırılmıştır. İçsel reaktivitesi oldukça yüksek olan A numunesi için modelin ısıl sürüklenme zamanı ve kendi kendine ısınma eğrisi tahmini, deneysel sonuçla oldukça uyumludur. Daha güvenli olarak tanımlanabilecek, düşük reaktiviteye ve yüksek nem içeriğine sahip B numunesinin davranışını ise model oldukça başarılı bir şekilde tahmin etmiştir. Bu tezde önerilen model ve yöntem kullanılarak, kömürün kendi kendine tutuşma yatkınlığının belirlenmesinde en etkili yöntem olarak görülen ve haftalarca sürebilen kuluçka testleri çok kısa bir süre içerisinde tamamlanarak, simüle edilen kömür için tam bir sıcaklık – zaman değişimi elde edilebilir.

Self-heating and ignition behavior of coal under mine conditions is a complex phenomenon and not yet fully understood. There are many parameters that effect this behavior. Intrinsic reactivity and moisture content are the most important factors. The aim of this study is to examine the self – heating and ignition of moist coal experimentally and theoretically. A new reaction kinetics expression has been proposed covering the whole low temperature region of coal self-heating. Also, a comprehensive evaporation/drying model has been developed and validated to express the moisture removal rate of coal during self-heating. By combining these two improvements for describing the self-heating process a more robust mathematical model has been developed. There are many experimental methods available to assess coal spontaneous combustion propensity. Incubation test method has been adopted for benchmarking throughout the coal industry for hazard management assessment in recent years. Incubation testing can measure the time taken to reach thermal runaway for coals across the rank spectrum. Since the coal is tested in its as-received moisture state the test enables the competing effects of coal oxidation and moisture removal to be measured, in other words the on-going heat balance between these two processes as a function of time and temperature. An experimental setup; consists of a perfectly insulated reaction vessel, an adiabatic oven, which is controlled automatically to maintain the oven temperature equal to that of coal in the reaction vessel, and a heat exchanger for preheating the gas to the oven temperature; has been constructed for his study. All heating elements, flow rates, valves are controlled with PLC and relevant data is uploaded to a computer for data logging. A heating control algorithm has been developed and maintains a ± 0.1oC adiabatic environment for critical low temperature oxidation phase. Self-heating of coal starts with adsorption of oxygen on the fresh coal surfaces at low temperatures. A decrease in oxidation/reaction rate is observed in the early stages of the process. Reactive sites, that adsorb oxygen, are responsible for this behavior. In weathered or pre-oxidized coals, oxygenated complexes deactivate these reactive sites. The main problem for determining the reaction rate of a moist coal is the complexity of representing active sites in the model. During the initial stages of oxidation, active sites bind the oxygen on the inner surfaces of the coal and oxygenated compounds form on the surfaces. These active sites become deactivated after prolonged contact with oxygen. This process has been represented as an ageing effect in the model. Moisture evaporation and removal from coal structure has a strong effect on self-heating behavior of moist coal. Latent heat of evaporation absorbs heat from coal and inhibits the temperature rise and delays or prevents the thermal runaway. Therefore it is crucial to analyze the evaporation thoroughly for a complete modelling of the self-heating of a moist coal. In previous studies, rate of evaporation of moisture has been determined by using equilibrium moisture content with a linear function of relative humidity of the surrounding air and coal moisture content. This expression of evaporation cannot be used in modeling self-heating incubation tests since the test is performed with dry oxygen and air flow rate and particle size have not been incorporated. In this study, a bound water resistance term has been introduced in the model for the first time by modifying the basic mass convection of moisture. A thermo-balance and a climatic chamber are used to validate the drying kinetics model. A series of experiments have been performed changing the temperature and flow rate with the same coal. In the first experiment, coal has been dried under 10 mL/min flow rate of nitrogen at 80oC. Weight change of coal has been continuously recorded with a 1-minute interval. This experiment has been used to derive σ, the characteristic evaporation constant, of the model. Temperature and gas flow rate are the key parameters for drying. Changing these parameters, two more series of experiments have been carried out with the same coal. It has been dried (at 90oC under 10 mL/min flow rate) and (at 80oC under 50 mL/min flow rate) to validate the temperature and flow rate dependent parameters of the model. Model predictions are found to be acceptable. R70 test is used to validate the reaction kinetics model for dried coal since there are no interactions other than oxygen – coal reactions. The model predicts the experimentally measured trend perfectly. Simulations have been carried out using an arbitrary-precision arithmetic algorithm in MATLAB for derived mathematical model. Model predictions of thermal runaway time and shape of the self-heating curve fit well with the experimental results for a highly reactive coal. However, the model predicts a moisture shoulder around 105oC, while experimental results show that this shoulder is around 95oC. When the coal temperature reaches this value, the sample begins to lose heat with increasing evaporation and the temperature stays at that level for almost a day. The moisture shoulder temperature is difficult to predict due to the highly nonlinear nature of kinetic rate terms. A slight change on the rates has significant impact on heat balance between oxidation and evaporation which is directly related to this temperature. For highly reactive, low rank coals, rate coefficients should be determined more precisely. To highlight this behavior, a less reactive high moisture containing coal sample has been investigated. The self-heating incubation test result records that temperature increases to a maximum of 63.9oC, then no thermal runaway occurs, even after 90 hours. The model predicts this behavior very well. This new model provides a better theoretical understanding of the self-heating incubation behaviour of moist coal. The first and main improvement in the model is related to the reaction kinetics of the moist coal oxidation. This new reaction kinetics model considers both incubation (slow self-heating with time) and the fast initial oxidation phases. Both these spontaneous combustion phases are controlled by the intrinsic coal reactivity and moisture heat balance. As such, moist coal reactivity depends on the moisture present in the coal (both in the pore spaces and on the internal pore surfaces) and more importantly it depends on the moisture removal/drying rate. Reactivity increases with the drying rate of moist coal. Hence, a fast dried coal sample has a higher reactivity than a slow dried coal sample. This implies realistic moist coal spontaneous combustion propensity can only be measured by tests conducted on the coal in its as-received moisture state. The other important improvement in the model concerns evaporation and moisture removal from the moist coal, which creates a heat loss mechanism that determines the moist coal self-heating behaviour. When the evaporation rate is too high during the incubation period, the moist coal cannot reach the thermal runaway state The new model produces accurate simulations of the self-heating behaviour of both dry and moist coal. It can be used to determine if thermal runaway is possible and in what timeframe it occurs. This has been validated from adiabatic testing results for different coals. An additional benefit of the new model is that it can be used to shorten the duration of present incubation testing, since these can often take several days or even weeks to reach completion. Once the self-heating of the sample reaches the moisture shoulder phase, the oven temperature can be set to maintain a slightly elevated temperature, thus shortening the drying time of this phase. The last section of the test through to thermal runaway can then be completed by switching the oven back to adiabatic tracking conditions. The results obtained from the first and last stage of the experiment can then be fed into the model to produce the entire time-temperature history of the coal self-heating.