Raghunathan Rengaswamy

Indian Institute of Technology Madras, Chennai, Tamil Nadu, India

Are you Raghunathan Rengaswamy?

Claim your profile

Publications (151)239.2 Total impact

  • Chris de Beer · Paul S. Barendse · Pragasen Pillay · Brian Bullecks · Raghunathan Rengaswamy
    [Show abstract] [Hide abstract]
    ABSTRACT: This paper presents the evaluation of a high-temperature proton exchange membrane (HTPEM) fuel cell for different degradation mechanisms using equivalent circuit analysis. Specific consideration is given to the variation of phosphoric acid content in the polybenzimidazole membrane and the effect on the equivalent circuit. The importance of the cell assembly and operating conditions on acid migration are discussed, and it is shown how it affects performance both in the short-term and the long-term operation of the HTPEM cell. A new method is developed, whereby acid leaching can be greatly accelerated in order to quantify performance loss. The change in system parameters as a function of membrane electrode assembly acid content is investigated using electrochemical impedance spectroscopy and compared with the changes that take place for catalyst degradation, CO poisoning, and reactant starvation. It is shown, by using the equivalent circuits, that the drop in performance relating to the individual degradation mechanisms in the cell can be investigated and isolated for fault classification in online diagnostic systems.
    IEEE Transactions on Industrial Electronics 08/2015; 62(8):1-1. DOI:10.1109/TIE.2015.2393292 · 6.50 Impact Factor
  • Chris de Beer · Paul S. Barendse · Pragasen Pillay · Brian Bullecks · Raghunathan Rengaswamy
    [Show abstract] [Hide abstract]
    ABSTRACT: High-temperature proton exchange membrane (HTPEM) fuel cells are able to withstand a substantial amount of CO poisoning while still maintaining stable output. High concentrations of CO will however degrade performance and can lead to instability. This paper presents a detailed study on the impact of CO poisoning on the performance of High-temperature PEM fuel cells and proposes a new method for mitigating the long term effects by using natural current profiles. A dedicated test setup was constructed to perform both steady state and dynamic analysis on the fuel cell under a wide range of operating conditions and variations in CO content. The drop in performance captured in the polarization curves is modeled using a simple circuit model with a dedicated fault element. The captured impedance spectra from the electrochemical impedance spectroscopy tests provide insight to the changes in the electrochemical circuit parameters that can be used to diagnose the extent of CO poisoning. Possible load control strategies that can reverse the CO poisoning is explored and the optimal profile is experimentally determined along with the long term effects.
    IEEE Transactions on Industry Applications 01/2015; 51(1):619-630. DOI:10.1109/TIA.2014.2328786 · 2.05 Impact Factor
  • Pratik Pednekar · Debangsu Bhattacharyya · Richard Turton · Raghunathan Rengaswamy
    [Show abstract] [Hide abstract]
    ABSTRACT: In slagging gasifiers, the molten slag gets deposited on the gasifier wall eventually causing refractory degradation. Replacement of the gasifier refractory is not only expensive, but it also leads to significant downtime. A computational model can help understand the impacts of various operating conditions on the refractory degradation and can be utilized for developing mitigation strategies. Therefore, the focus of this work is to develop rigorous, dynamic, first-principles models integrated with comprehensive mechanistic models that can be used for estimating the extent of degradation of the gasifier wall at any location at any point of time. Traditionally while modeling the entrained-flow gasifiers, it is assumed that as the char reacts, the ash remains attached to it leading to higher mass transfer resistances. Therefore a shrinking core model is used. However, the gasifier operating temperature is much higher than the melting temperature of the ash and therefore, it is likely that molten droplets of slag will form on the char surface. Due to strong detachment forces on the surface of a reacting char particle, these droplets may separate and some of them may eventually get deposited on the wall. To capture this phenomenon, a shrinking particle model is developed where the size of the char particles keeps decreasing due to the heterogeneous reactions leading to formation of molten slag droplets. These slag droplets remain attached to the char surface till the detachment forces exceed the adhesive forces. These mechanisms are modeled by combining the continuum model of the gasifier with a discrete particle model that accounts for the size of the slag droplets attached to the char particle and the population and size of the slag droplets in the bulk. A comprehensive transport model is developed for calculating the flux of the slag droplets towards the wall. The model estimates the numbers and sizes of the slag droplets that hit the refractory wall at a particular location. In addition, a model for the slag layer is developed to calculate the thickness of the slag layer along the wall and the temperature profile across it. Finally, a model is developed to capture penetration of the molten slag inside the refractory and the resulting degradation of the wall due to tensile and compressive forces. Both the continuum and discrete particle models are developed in Aspen Custom Modeler® (ACM) environment. The gasifier dynamics are studied by simulating a number of disturbances. In addition to results from these studies, the presentation will also include results that show the effects of gasifier operating conditions on population and sizes of the slag droplets, slag layer thickness, wall temperature profile, and refractory degradation.
    14 AIChE Annual Meeting; 11/2014
  • Parham Mobed · Jeevan Maddala · Debangsu Bhattacharyya · Raghunathan Rengaswamy
    [Show abstract] [Hide abstract]
    ABSTRACT: Integrated gasification combined cycle (IGCC) with CO2 capture is an attracting technology for power generation due to its high efficiency and near-zero emission. A particular unit that is crucial to the performance of IGCC is the water gas shift reactor (WGSR), which is responsible for adjusting the syngas H2/CO ratio to meet power production and carbon capture targets. The efficacy of the WGSR is greatly affected by number of disturbances, such as upstream feed disturbance, and faults, such as catalyst deactivation due to poisoning and fouling. An early detection of disturbances and faults can facilitate the preventive actions and ensure safe and efficient operation of the reactor. However, faults and disturbances are not essentially observable by measurements and require a state estimation of the internal state of the reactor for them to be detected. This can be done by using a first-principle model of the reactor and series of measurements distributed throughout the reactor. It is therefore necessary to find the optimal number and location of the sensors for efficient estimation of faults and disturbances. A high-fidelity model of the WGS reactor was developed in our previous work. The model is presented by a system of non-linear differential and algebraic equations (DAE). The discretized equations form a space of grid-points that are available for state measurement. An extended Kalman filter was implemented to estimate the faults using the measurements chosen from the space. A genetic algorithm is then employed to search the space for obtaining an optimal location of the sensors. However, the approach is inefficient due to large search space and computational burden for non-linear state estimation. To address this, a reduced order model of the WGSR is developed by linearization of the non-linear model. The genetic algorithm uses the linearized model and a linear Kalman filter to find the optimal location of the sensors. This approach significantly reduces the computational work compared to our previous method.
    14 AIChE Annual Meeting; 11/2014
  • Parham Mobed · Jeevan Maddala · Debangsu Bhattacharyya · Raghunathan Rengaswamy
    [Show abstract] [Hide abstract]
    ABSTRACT: Tight carbon emission standards on power generation from fossil fuels have made researchers focus on plants with near-zero emissions. Integrated gasification combined cycle (IGCC) power plants with CO2 capture are promising technologies for safe and clean power generation that enforce the environmental emission standards by treating the fuel gas in the acid gas removal (AGR) unit. Satisfying the environmental constraints depends crucially on the performance of AGR unit. Any faults that occur in the AGR unit can drive the process away from its nominal condition and may lead to violation of environmental constraints and hazardous consequences. An early detection and identification of the faults facilitate preventive actions for safe and optimal operations. Any abnormalities affect several process variables throughout the process which requires measurements of variables for symptoms to be observed. In order to detect and diagnose faults, crucial variables must be identified based on their economical and practical feasibility. An algorithmic approach for identifying the optimal number, type and location of the sensors for fault detection and diagnosis is useful, particularly for the large-scale, emerging fossil power plants with CO2 capture. The usefulness of magnitude ratio (MR) algorithm developed in our previous work[1], which enhances the diagnosis capability of signed directed graph (SDG) sensor placement algorithm, is demonstrated by implementing on three basic case studies. The algorithm uses all the variables from the process and assumes the magnitude ratio of each pair as a pseudo-sensor. Similar to SDG, pseudo-sensors have discrete states and are treated by symmetric difference operator in sensor network design framework. When used in SDG framework, the algorithm retains the properties of SDG and manages to add more information for improving diagnosis capability. The algorithm is implemented on the SELEXOL-based AGR unit to identify an optimal sensor network. The enhanced diagnosis capability of the MR algorithm observed in AGR unit is promising for further implementation of the algorithm in large and complex power plants. Reference: [1] P Mobed, J Maddala, D Bhattacharyya, R Rengaswamy, "On the Use of Magnitude Ratio in Sensor Placement Algorithms for Fault Detection and Diagnosis in Complex Energy Processes", 57th Annual ISA POWID Symposium, Scottsdale, Arizona, USA, 2014.
    14 AIChE Annual Meeting; 11/2014
  • Pratik Pednekar · Debangsu Bhattacharyya · Richard Turton · Raghunathan Rengaswamy
    [Show abstract] [Hide abstract]
    ABSTRACT: Sensors can be useful in the early detection of faults in the Integrated Gasification Combined Cycle (IGCC) system thereby preventing damage to life and property and ensuring efficient energy production. Fault such as refractory degradation in the gasifier cannot be measured directly with the current state of the art in measurement due to the extreme conditions within the gasifier. A sensor or a set of sensors can be placed at some other location so that it’s (their) response(s) can be used to monitor the condition of the refractory wall. A component-level model that is high-fidelity partial differential algebraic equation (PDAE)-based model is required for developing the quantitative model-based approach for sensor placement for component condition monitoring. The gasifier is the heart of the IGCC power plant. Due to its operation at very high temperatures, the ash content in the coal melts to form slag. This slag gets deposited onto the gasifier wall and penetrates into the refractory bricks leading to refractory degradation. Traditionally, in modeling a slagging entrained flow gasifier, slag deposition is assumed to happen only when the char particle itself impacts the gasifier wall. However, it is reported that the ash content can melt while in the bulk phase itself. The slag formed can separate from the char particle due to the detachment forces in a process called shedding. These slag droplets can then themselves deposit onto the flowing slag layer on the gasifier wall. These processes are modeled by combining the continuum model of the gasifier and discrete particle model to account for the size and number of slag droplets. A comprehensive transport model is developed to calculate the flux of the slag droplets to the wall. This is required in order to model the flowing slag layer on the gasifier wall which in-turn will calculate the thickness of the slag layer and the temperature profile across it. Finally, slag penetration into the refractory and refractory degradation due to the mechanisms of tensile and compressive spalling is modeled. Using the PDAE based model and measurements from the gasifier, the sensor network will be able to estimate the condition of the refractory in real time.
    14 AIChE Annual Meeting; 11/2014
  • [Show abstract] [Hide abstract]
    ABSTRACT: Integrated gasification combined cycle (IGCC) plants with CO2 capture have strong potential in the future carbon-constrained world. In these plants, the CO2 and COS content of the syngas at the inlet of the acid gas removal process should be within certain limits in order to satisfy the environmental limits on sulfur and CO2 emissions. To satisfy these constraints, the syngas from the gasification process is passed through water gas shift reactor(s). A premium is placed on sulfur-tolerant catalysts since the syngas may contain COS and H2S. In comparison to the sweet-shift process, the sour-shift process results in higher overall efficiency because of the higher temperature of the feed syngas and requirement of less additional steam for the shift reactions. The optimal operating conditions and the dimensions of the sour shift reactors can be obtained by considering the effect of a number of key variables. With this motivation, a 1-D mathematical model of a sour water gas shift (WGS) reactor has been developed by considering mass, momentum, and energy conservation equations. The experimental data available in the open literature are reconciled for measurement errors after gross errors are removed and then used to obtain the rate parameters. Subsequently, the developed model is used to study the performance of the WGS reactor as part of an IGCC plant with CO2 capture. The results presented in this paper are very useful in designing, analyzing, and operating the sour water gas shift reactors.
    Industrial & Engineering Chemistry Research 09/2014; 53(51):19855-19869. DOI:10.1021/ie500739h · 2.24 Impact Factor
  • Tim Spinner · Babji Srinivasan · Raghunathan Rengaswamy
    [Show abstract] [Hide abstract]
    ABSTRACT: This work presents a new look at the existing data-based and non-intrusive PI (proportional-integral) controller tuning assessment methods for SISO (single-input single-output) systems under regulatory control. Poorly tuned controllers are a major contributor to performance deterioration in process industries both directly and indirectly, as in the case of actuator cycling and eventual failure due to aggressive tuning. In this paper, an extensive review and classification of performance assessment and automated retuning algorithms, both classical and recent is provided. A subset of more recent algorithms that rely upon classification of poor tuning into the general categories of sluggish tuning and aggressive tuning are compared by their diagnostic performance. The Hurst exponent is introduced as a method for diagnosis of sluggish and aggressive control loop tuning. Also, a framework for more rigorous definitions than previously available of the terms “sluggish tuning” and “aggressive tuning” are provided herein. The performance of several tuning diagnosis methods are compared, and new algorithms for using these tuning diagnosis methods for iterative retuning of PI controllers are proposed and investigated using simulation studies. The results of these latter studies highlight the possible problem of loop instability when retuning based upon the diagnoses provided by data-based measures.
    Control Engineering Practice 08/2014; 29:23–41. DOI:10.1016/j.conengprac.2014.03.005 · 1.91 Impact Factor
  • Source
    Danny Raj Masila · Raghunathan Rengaswamy
    [Show abstract] [Hide abstract]
    ABSTRACT: In a microchannel device with two phase flow, the non-zero interfacial tension allows us to compartmentalize one of the fluids as the drop while the other flows as a continuous phase. The non-linear interactions between the drops inside the microchannel results in complex emergent behavior. A simple 1D loop device (Figure 1 a) can convert an equi-spaced train of drops into an exit train that can be periodic, aperiodic or chaotic [1]. Drops can be actively synchronized in a 1D ladder device (Figure 1 b) [2]. Drops entering a diverging-converging channel (Figure 1 c) can dynamically self-organize [3]. Coalescence in concentrated emulsions could propagate throughout the channel resulting in local phase inversion (Figure 1 e) [4]. Understanding the collective behavior of drops inside these channels using simple models becomes crucial in device design. A ‘compact network model’ was developed for understanding drop motion in 1D microchannels [5,6]. Our group was able to use this simple model in a model predictive controller [7] and a genetic algorithm formulation [8] to actively and passively manipulate drops in 1D channels respectively. Recently we proposed the ‘interacting drop-traffic models’, a set of phenomenological models to characterize the different interactions between the drops inside a microchannel. These models upon incorporation inside a multi-agent framework was able to predict the emergent ordered patterns (Figure 1 d) inside the channel [9]. Now a pair of drops coalescing in these self-organized patterns can trigger an avalanche of coalescaence events (Figure 1 f). We attempt to model this propagated coalescence through a stochastic framework. The simultaneous self-organization and propagation inside the channel exhibits characteristics similar to a self-organized critical process [10].
    DDAP 08: Dynamic Days Asia Pacific 08, IIT-Madras-IMSC, Chennai, India; 07/2014
  • Babji Srinivasan · Ulaganathan Nallasivam · Raghunathan Rengaswamy
    [Show abstract] [Hide abstract]
    ABSTRACT: In industrial plants with non-oscillatory set points, oscillation detection and diagnosis is a key step to improve plant performance and safety. Oscillations in linear closed loop systems can occur due to one or more of the following reasons: (i) changes in process/controller settings, (ii) stiction in control valves, (iii) external oscillatory disturbances, (iv) quantization effects and, (v) presence of saturation and hysteresis in closed loop systems. Though there are techniques to address oscillation diagnosis problem, there are gray areas such as the identification of multiple sources that cause oscillations in the process output. In this work, this problem is addressed through the development of an algorithm to identify multiple sources of oscillations in Single Input Single Output (SISO) loops. Further, an integrated approach to diagnose both single/multiple root causes in SISO loops is presented. Simulation and industrial case studies are provided to show the applicability of the proposed algorithms.
    Chemical Engineering Research and Design 07/2014; 93. DOI:10.1016/j.cherd.2014.07.004 · 2.28 Impact Factor
  • Danny Raj Masila · Raghunathan Rengaswamy
    [Show abstract] [Hide abstract]
    ABSTRACT: Recently we proposed a simple, computationally inexpensive multi-agent simulation strategy to understand drop movement in a microchannel with Hele-Shaw flow geometry (2D). This has helped us understand the collective behavior of the drops in these systems. However, transitioning these systems to practice would require that strategies are developed to combat the inherent fluctuations that are a part of any physical system. In the present work we utilize the simulation strategy for understanding the impact of design choices in minimizing the effect of uncertainties, and the sensitivity of the device behavior to operational variables that can potentially be used as controlled variables. We study a control metric, time of contact, which is defined as the time spent by a drop in the microchannel when its distance from its neighboring drops is less than some critical value. By simplifying the model further we are able to analyze the contact time dynamics as a function of geometric parameters and operating conditions analytically. We also carry out the complete numerical simulation of our models to understand active control of drop clusters.
    European Control Conference ECC 14, Strasbourg, France; 06/2014
  • Laya Das · Babji Srinivasan · Raghunathan Rengaswamy
    [Show abstract] [Hide abstract]
    ABSTRACT: A new technique is developed for assessing the performance of linear and nonlinear Kalman filter based state estimators. The proposed metric will indicate the performance of these state estimators which will be primarily influenced by: (i) difference between the model dynamics and process dynamics and, (ii) various approximations of the nonlinear plant dynamics used in nonlinear Kalman filters. Currently, there exists no such quantification method to analyze the performance of linear and nonlinear Kalman filters, a key requirement for improvement and a practical benchmark for comparison of these state estimation algorithms. The proposed technique uses the generalized Hurst exponent of the prediction errors (difference in measured output and a posteriori estimates) obtained from the state estimators to quantify the performance. This technique could be implemented on-line as it requires only plant operating data and the predicted outputs (from the linear and nonlinear Kalman filters) to assess the performance. Several simulation studies demonstrate the applicability of the proposed performance metric to both linear and non-linear Kalman filters.
    2014 American Control Conference - ACC 2014; 06/2014
  • Source
    K. Villez · R. Rengaswamy
  • Source
    K. Villez · R. Rengaswamy
  • Jeevan Maddala · Siva A. Vanapalli · Raghunathan Rengaswamy
    [Show abstract] [Hide abstract]
    ABSTRACT: Droplets moving in a microfluidic loop device exhibit both periodic and chaotic behaviors based on the inlet droplet spacing. We observe that the periodic behavior is an outcome of carrier phase mass conservation principle, which translates into a droplet spacing quantization rule. This rule implies that the summation of exit spacing is equal to an integral multiple of inlet spacing. This principle also enables identification of periodicity in experimental systems with input scatter. We find that the origin of chaotic behavior is through intermittency, which arises when drops enter and leave the junctions at the same time. We derive an analytical expression to estimate the occurrence of these chaotic regions as a function of system parameters. We provide experimental, simulation, and analytical results to validate the origin of periodic and chaotic behavior.
    Physical Review E 01/2014; 89(2). DOI:10.1103/PhysRevE.89.023015 · 2.33 Impact Factor
  • Jeevan Maddala · Raghunathan Rengaswamy
    [Show abstract] [Hide abstract]
    ABSTRACT: Accelerated progress in the use of droplet-based microfluidics for high throughput screening and biochemical analysis will require development of devices that are robust to experimental uncertainties and which offer multiple functionalities. Achieving precise functionalities in microfluidic devices is challenging because droplets exhibit complex dynamic behavior in these devices due to hydrodynamic interactions and discontinuities that are a result of discrete decision-making at junctions. For example, even a simple loop device can show transitions from periodic to aperiodic/chaotic behavior based on input conditions. Hence, rational design frameworks that handle this complexity are required to move this field from labs to industrial practice. Two main challenges that need to be confronted in the realization of such a rational design framework are: (i) computational science related to rapid simulation of very large networks; development of predictive models that will form the basis for characterizing droplet motion through interconnected and intricate large-scale networks, and (ii) conceptualization of a design approach that is generic in nature and not very narrowly defined limiting its application potential. In this paper, we develop a GA approach for the design of ladder networks that are used to control the relative droplet distance at the exit. Through several case studies, the potential of the proposed GA approach in designing exquisite ladder structures for multiple functions is demonstrated. A recently proposed network model is used as the basis for all the computational studies reported in this paper.
    Computers & Chemical Engineering 01/2014; 60:413–425. DOI:10.1016/j.compchemeng.2013.09.009 · 2.45 Impact Factor
  • Source
    Raj M Danny · Raghunathan Rengaswamy
    [Show abstract] [Hide abstract]
    ABSTRACT: Drops can be used as carriers of reactants and there could be transfer of mass between the drops when they come close to each other, allowing one to use this system as a micro-reactor system. In this paper, we use the simple models that were developed in a recent work (Danny and Rengaswamy, 2013) to explain the different forces on the drops inside a 2-D microchannel, in a multi-agent based simulation to study the performance of a diverging-converging microchannel as a drop-drop contactor. Our models are applicable to a wide range of 2-D geometries that are symmetric about the horizontal axis. A geometry similar to the one experimentally studied by (Jose and Cubaud 2011) is chosen to demonstrate the technique. In this system the initial spacing between the drops prior to entry into the microchannel affects the configuration of the drops inside the microchannel resulting in layers of drops. This motivates the investigation of the different patterns formed inside the microchannel as a function of the initial spacing between the drops prior to their entry into the channel. To study the efficiency of drop-drop contactor, the case where there are two different reactants in two different drops A and B is considered. One would simply expect an entry sequence of ABAB to get maximum contact between drop-A and drop-B. After a study of the patterns formed by drops inside the microchannel for a given entry sequences, it is found that ABAB is a favorable sequence only for a very small class of problems and the least preferred for the rest. A metric called the contact efficiency (CE) is defined which in some sense gives the number of drops A around B and vice versa, to quantify the efficiency of the pattern formed, to compare the effect of different input sequences.
    AiChe Annual meet 2013, San Francisco, California, USA; 11/2013
  • Pratik Pednekar · Debangsu Bhattacharyya · Richard Turton · Raghunathan Rengaswamy
    [Show abstract] [Hide abstract]
    ABSTRACT: Modeling Refractory Degradation due to Slag Penetration in an Entrained-Flow Gasifier Pratik Pednekar, PhD Student, West Virginia University Department of Chem. Eng., West Virginia University, Morgantown WV 26506 ppedneke@mix.wvu.edu Debangsu Bhattacharyya, Assoc. Professor, West Virginia University Department of Chem. Eng., West Virginia University, Morgantown WV 26506 Debangsu.Bhattacharyya@mail.wvu.edu Tel: 3042939335, Fax: 3042934139 Richard Turton, Professor, West Virginia University Department of Chemical Engineering, WVU, Morgantown, WV 26506 Richard.Turton@mail.wvu.edu Tel: 3042939364, Fax 3042934139 Raghunathan Rengaswamy, Professor, Texas Tech University Department of Chemical Engineering, Texas Tech University, Lubbock, TX 26507 Raghu.Rengasamy@ttu.edu Tel: 8067421765, Fax 3042850903 Abstract In slagging gasifiers, molten slag flows down the inner refractory wall. The slag can penetrate into the refractory causing thinning and spalling of the refractory lining due to the build-up of stress. Replacement of the refractory lining is required every 1-2 years and is very expensive. This also leads to significant down-time lowering the overall availability of gasifier-based power plants. Due to the harsh operating condition inside a slagging gasifier, an in-situ measurement of the refractory degradation is not possible with the current state of technology. Therefore, a mathematical model may be used to develop a better understanding of the effects of various operating conditions and physicochemical properties on the degradation characteristics. Such a model may eventually lead to development of better monitoring and prevention techniques. With this motivation, first a slag submodel was developed to obtain the temperature profile in the wall and the thickness profile of the slag along the gasifier wall. A mechanistic model was developed by considering slag formation and detachment from the char particles followed by transport and deposition of the slag droplets onto the refractory wall. A model for the slag layer was then developed by considering mass, momentum, and energy conservation equations. Due to the temperature gradient that exists in the molten slag layer, a solid slag layer is formed between the refractory and the molten slag layer. A model of the solid slag layer was also developed using energy balance equations. Two phenomenological models were then developed for refractory degradation - one based on the tensile stress and the other based on the compressive force developed due to thermal strains and shrinkage/swelling due to slag penetration. The slag sub-model was integrated with a 1D steady-state model of a single-stage, downward-firing, oxygen-blown, slurry-fed, entrained-flow gasifier developed previously at WVU in the Aspen Custom Modeler (ACM) environment. The gasifier model included mass, momentum and energy balance equations for solid and gas phases. The model also included a number of heterogeneous and homogeneous chemical reactions along with devolatilization and evaporation of the slurry feed. The presentation will include results that show the effects of gasifier operating conditions on slag layer thickness, wall temperature profile, and refractory degradation.
    13 AIChE Annual Meeting; 11/2013
  • [Show abstract] [Hide abstract]
    ABSTRACT: Coal-fed integrated gasification combined cycle (IGCC) is a promising technology for electricity generation with zero emissions. In the catalytic water gas shift reactors (WGSRs), as part of an IGCC plant with carbon dioxide capture, the carbon monoxide present in the syngas reacts with steam to generate carbon dioxide and hydrogen. The carbon dioxide is captured in an acid gas removal unit and the hydrogen-rich syngas is used as a fuel for power generation. The efficiency and performance of the IGCC plant is significantly affected by the WGSRs, which in turn, are affected by a number of disturbances and faults. For instance, disturbances in flow, temperature or molar compositions of the feed affect the conversion and, consequently the efficiency of the reactor. Additionally, the catalyst inside the WGSR is vulnerable to faults due to poisoning, fouling, or thermal cycling resulting in deactivation, porosity reduction, and/or surface area reduction. Therefore, an early detection of disturbances and faults in the WGSRs can ensure safe and efficient operation of the reactor. A first-principles model of the WGSR can be used to predict the states of the reactors and detect the faults. Since process models are never completely accurate and the external disturbances are not predictable, a series of measurements, such as measurements from temperature, pressure and concentration sensors, along with the process model is helpful in estimating the unknown states of the reactor. However, the challenging task is to find the optimal number and location of these sensors for estimation of faults in the presence of disturbances and uncertainties in the process model and measurements. A first-principles model of the WGS reactor has been developed in our previous work. Since the resulting DAE system is non-linear, an extended Kalman filter (EKF) is tailored for estimation in DAE systems. Possible faults, which are identified with prior process knowledge, are incorporated as states and estimated along with other states of the system. An accurate estimation of the states can help in identifying the disturbances and detecting the faults. Therefore, minimizing the error in the estimated states requires finding the best measurement model that contains the optimal number and location of the sensors. The measurement model is a discrete set of binary values for all measureable states, representing 1 if a sensor is placed and 0 if no sensor is placed for the specific discrete variable. In order to efficiently detect all the faults, a combinatorial optimization problem is solved to find the best measurement model while minimizing the error between the actual and estimated states of the reactor. In this work, a genetic algorithm which can handle discrete optimization problems is proposed, where at each generation, a population of measurement models is used in the EKF with genetic operations performed between each generations, until the best solution is found. In this presentation, we will present the SP algorithm and the optimal measurement model that can detect faults with reasonable accuracy in the presence of disturbances, measurement noises and uncertainties in the process model.
    13 AIChE Annual Meeting; 11/2013
  • M. R. Islam · Raghunathan Rengaswamy · Debangsu Bhattacharyya · Richard Turton
    [Show abstract] [Hide abstract]
    ABSTRACT: Early detection of faults in process plants can help in improving plant safety, achieving higher efficiency and attaining longer equipment life. One or more process variables may get affected due to occurrence of a fault in process plants. However, measurement of every variable is neither economically and practically feasible nor necessary. An optimal set of sensors can be selected to diagnose the faults. However, placing an optimal number of sensors for detection and identification of faults becomes challenging for large and complicated processes. Most of the sensor placement (SP) problems for large processes have been solved in the existing literature by using qualitative models of the plant because of easy conversion of the cause-effect (CE) relations into graphical and mathematical representations. Among various qualitative model-based SP approaches, the directed graph (DG) and signed directed graph (SDG) based approaches are widely used. In the DG and SDG based approaches, variables are chosen for fault observabilty and resolution from the sets of variables that respond due to the faults. However, for fault resolution, variables are selected from the sets of responding variables for a pair of faults. In the DG based approach, when two or more different faults yield same set of responding variables, those faults cannot be distinguished from each other. Even the SDG based approach cannot distinguish these faults if the responding variables deviate in the same direction in each set. To improve the effectiveness of the sensor network in resolving faults and to obtain a measurement network with fewer sensors, a SP algorithm is developed by including the magnitude of deviation of the responding variables and the propagation time of faults through the process variables. Even though two different faults may yield the same set of responding variables with the same directionality for each variable, the magnitude of deviation can be very different for each fault. The pair-wise ratio of the magnitude of deviation of the responding variables is found to be successful in distinguishing between faults that could not be resolved with the DG or SDG based approaches. In addition different process variables can respond to the same fault at different times. For a particular fault, one variable may respond earlier than another variable. Selection of this pair of variables for the sensor network helps to identify that particular fault. In this work, a sensor placement algorithm has been developed by using magnitude ratios and fault propagation times for fault resolution. A set cover problem is formulated subject to fault observability and resolution. The proposed algorithm is employed to find optimal sensor network for a simple CSTR as well as the Tennessee Eastman process.
    13 AIChE Annual Meeting; 11/2013

Publication Stats

3k Citations
239.20 Total Impact Points

Institutions

  • 2013–2014
    • Indian Institute of Technology Madras
      • Department of Chemical Engineering
      Chennai, Tamil Nadu, India
  • 2009–2014
    • Texas Tech University
      • Department of Chemical Engineering
      Lubbock, Texas, United States
  • 2001–2012
    • Clarkson University
      • Department of Chemical and Biomolecular Engineering Coulter School
      Potsdam, New York, United States
  • 2001–2011
    • Purdue University
      • School of Chemical Engineering
      West Lafayette, Indiana, United States
  • 1999–2002
    • Indian Institute of Technology Bombay
      • Department of Chemical Engineering
      Mumbai, Maharashtra, India