Data processing: measuring – calibrating – or testing – Measurement system – Performance or efficiency evaluation
Reexamination Certificate
2003-08-13
2004-11-30
Hoff, Marc S. (Department: 2857)
Data processing: measuring, calibrating, or testing
Measurement system
Performance or efficiency evaluation
C702S182000, C702S183000, C702S184000, C702S189000, C702S190000
Reexamination Certificate
active
06826513
ABSTRACT:
FIELD OF THE INVENTION
The present invention relates to a method and an apparatus for online detection of safe operation and advance detection of unsafe operation of a system or process in the presence of noise in sensor measurements and/or fluctuations in variables measured. More particularly, the present invention provides a method and an apparatus capable of providing advance warnings of a system or process becoming unsafe during operation. Still more particularly, the present invention relates to a method and an apparatus that uses complex
oisy sensor measurements made available from sensors which monitor temperature, viscosity, thermal conductivity, chemical species concentrations, pressure or flow signals as a time-series from a batch, continuous stirred tank, fixed-bed, biochemical, polymerization, fluidized-bed, catalytic
oncatalytic reactors, multiphase, flow or other physical systems operating under varied conditions for advance online identification of normal/safe operation along with detection of abnormal/unsafe operation.
BACKGROUND OF THE INVENTION PRIOR ART REFERENCES
Detection and advance warning of unsafe operation of a system in presence of commonly encountered measurement noise causing fluctuations in monitored sensor signals of either temperature, viscosity, thermal conductivity, mass diffusivity, chemical species concentrations, pressure or flow as time-series from batch, continuous stirred tank, fixed-bed, biochemical, polymerization, fluidized-bed, catalytic
oncatalytic reactors, multiphase, flow or other physical systems is of great importance in process applications. Online identification of normal safe process operation and detection of abnormal unsafe process operation in the presence of measurement noise from monitored digital sequences monitored from sensor apparatus as a time-series still evades satisfactory solution, despite efforts made over the years. The problem is particularly important in the context of chemical reactors where small changes in operating conditions can lead to unsafe process operating conditions resulting in risk to personnel safety, infrastructure, environment and process economics. It is important, therefore, to detect in advance the conditions when the system is initiated into an unsafe operating mode by an automated online procedure that uses time-series measurements that monitor process variable behavior. It is also important that this identification of safe process operation and advance detection of unsafe process operation be obtained when the fluctuations in the monitored signals arise due to process complexities on real time basis so that corrective actions can be taken to avoid the loss to life, property, environment, economics well in advance.
Unsafe process operating conditions in reactors arise when a small change in an operating variable like feed or coolant temperature, flow rate, concentrations of species, viscosity, thermal conductivity, mass diffusivity bring about a drastic change in variable values whereby the system becomes uncontrollable. For example, in the case of temperature, a small increase can induce a large increase in the reaction rates due to the exponential dependency of rate constants on temperature. The increased reaction rates further increases the temperature and this results in the process having the rate of heat generation far exceeding the rate of heat removal by the cooling equipment. Unsafe process operating conditions can damage the reactor vessel, can cause personnel risks, catalyst deactivation, choking, hotspot formation, environmental degradation, economic loss and it is important to a priori detect in advance the conditions which lead to this undesirable behavior so that they may be avoided. Studies related to the safety of process operation may be generally classified in two categories, viz., offline and online.
Offline methods analyze mathematical models of the reactor behavior for specifically chosen reaction systems with the aim of demarcating safe and unsafe regions of operation with respect to the values of the reactor control parameters (Van Welsenaere, R. J. and Froment, G. F., “Parametric Sensitivity and Runaway in Fixed-bed Catalytic Reactors”, Vol. 25 Chem. Engg. Sci. 1503 [1970]; Rajadhyaksha, R. A., Vasudeva, K. and Doraiswamy, L. K., “Parametric Sensitivity in Fixed-bed Reactors”, Vol. 30 Chem. Engg. Sci. 1399 [1975]; Morbidelli, M. and Varma, A., “Parametric Sensitivity and Runaway in Fixed-bed Catalytic Reactors”, Vol. 41 No. 4 Chem. Engg. Sci. 1063 [1986]; Balakotaiah, V., Kodra, D. and Nguyen, D., “Runaway Limits for Homogeneous and Catalytic Reactors”, Vol. 50 No. 7 Chem. Engg. Sci. 1149 [1995]; Heiszwolf, J. J., “Thermal Stability of Reacting Systems in Batch and Continuous Stirred Tank Reactors”, Ph. D. Thesis, University of Amsterdam [1998]). Generally, offline methods study the steady-state behavior and not the dynamic behavior to predict safe conditions for process operation. The safe and unsafe regions in relevant process parameters are marked on ready-made charts and tables for look-up. Because the mathematical model of the system is used in offline methods the criteria that are developed to test the safety have a serious drawback in the sense that they are conservative and do not apply in a generalized fashion. This is especially true for complex systems which are commonly encountered in many chemical, catalytic, polymerization, combustion, cracking, multiphase, flow and other physical systems.
On the other hand, few attempts have been made in developing online criteria using time-series signals from processes for the identification of safe operation and detection of unsafe operating conditions (Hub, L. and Jones, J. D., “Early Online Detection of Exothermic Reactions”, Vol. 5 Plant/Oper. Prog. 221 [1986]). Thus for instance, for exothermic reactions taking place in a reactor, temperature measurements by thermocouples, thermometers, thermistors, are usually available as monitored data of the temperature variable in time. The first derivative and higher-order derivatives of this monitored process variable turning positive are a signature of unsafe operation. But, a very significant drawback of these methods is that computation of derivatives is error-prone because of the commonly encountered fluctuating nature of the measurements especially when the measurement is noisy or the process is associated with fast time scales. Thus, the frequency of monitoring and the sample size used for calculation can have significant bearing on detecting unsafe conditions (Iserman, R., “Process Fault Detection Based on Modeling and Estimation Methods—A Survey”, Vol. 20 Automatica 387 [1984]). A possible way is to calculate derivatives by processing the signal with filters. Linear filters when applied to signals obtained from processes following nonlinear mechanisms is again error-prone because the true signal content may be inadvertently filtered. In principle, however, online methodologies have the advantage that they have generalization capability because the methods are applicable even when the mechanistic nature of the process or its mathematical model is unknown. Developing rigorous online methods when adequate modeling information is not known, but, takes into consideration the nonlinear properties of the process implicitly embedded in the data is important. Moreover, these methods would be most useful if they can also handle the presence of measurement noise while analyzing for safety. The advantages of developing online methods and apparatus for these aims would identify conditions when unsafe operation alarms are triggered so that other corrective measures are implemented on the process. On the other hand, as much as detection of unsafe conditions is important, the reverse situation of false alarms being raised is also deleterious to process operation and must be avoided as it affects process economics due to unnecessary shutdowns. Robustness in the method and apparatus for detection of unsa
Ghosh Anandamohan
Kulkarni Bhaskar Dattatraya
Kumar Viruthiamparambath Ravi
Council of Scientific & Industrial Research
Hoff Marc S.
Tsai Carol S W
LandOfFree
Method and apparatus for online identification of safe... does not yet have a rating. At this time, there are no reviews or comments for this patent.
If you have personal experience with Method and apparatus for online identification of safe..., we encourage you to share that experience with our LandOfFree.com community. Your opinion is very important and Method and apparatus for online identification of safe... will most certainly appreciate the feedback.
Profile ID: LFUS-PAI-O-3352855