Reactor upsets

Parameters such as feed rate, catalyst bed temperature, and reaction pressure were optimized by use of the temporary on-line LC installation. Reactor upsets could also be monitored. Figure 10 demonstrates how continuous monitoring can aid in detection of an upset. Due to a problem with a level control valve, the reactor filled with liquid, preventing the reaction... 

In 1990 the US Army Research Office organized the NATO Advanced Research Workshop, Destruction of Toxic Molecules in Supercritical Water . This workshop developed specific recommendations for research to provide fundamental understanding of the chemistry of hazardous materials in SCWO. Understanding the processes enables rational design of reactors, setting of optimum operation parameters, prediction of reactor materials lifetimes, and anticipation of reactor upsets. Some of the participants summarized that NATO workshop in an article for the American Chemical Society that brought the basic concepts of supercritical water as a reaction medium to a wide audience [11]. 

When low boiling ingredients such as ethylene glycol are used, a special provision in the form of a partial condenser is needed to return them to the reactor. Otherwise, not only is the balance of the reactants upset and the raw material cost of the resin increased, but also they become part of the pollutant in the waste water and incur additional water treatment costs. Usually, a vertical reflux condenser or a packed column is used as the partial condenser, which is installed between the reactor and the overhead total condenser, as shown in Figure 3. The temperature in the partial condenser is monitored and maintained to effect a fractionation between water, which is to pass through, and the glycol or other materials, which are to be condensed and returned to the reactor. If the fractionation is poor, and water vapor is also condensed and returned, the reaction is retarded and there is a loss of productivity. As the reaction proceeds toward completion, water evolution slows down, and most of the glycol has combined into the resin stmcture. The temperature in the partial condenser may then be raised to faciUtate the removal of water vapor. 

Distributors in industrial units typically have large numbers of injection points of quite diverse design characteristics, some of which are depicted in Eigure 16 for fluidized-bed appHcations. Flow variations through these parallel paths can lead to poor flow distributions within a reactor, thus reducing product yields and selectivity. In some circumstances, undesirable side products can foul portions of the distributor and further upset flow patterns. Where this is important, or where the possibiHties and consequences are insufficiently understood and independent means caimot be employed to assure adequate distribution, the pilot plant must be sized to accommodate such a distributor. Spacing should be comparable to those distributors that are anticipated to be... 

ADI Process. The ADI is a low rate anaerobic process which is operated ia a reactor resembling a covered football field. Because of the low rate, it is less susceptible to upset compared to the high rate processes. Its disadvantage is the large land area requirement. 

This section illustrates by way of example, the application of simphfied dispersion estimates to assessing a catastrophic venting operation. In this example, an analysis was performed to predict the fate of air pollutants, specifically vinyl chloride monomer (VCM), originating from an episode type upset (reactor blow) condition from a reaction vessel. 

Some batch reactions have the potential for very high energy levels. If all the reactants (and sometimes catalysts) are put into a kettle before the reaction is initiated, some exothermic reactions may result in a runaway. The use of continuous or semi-batch reactors to limit the energy present and to reduce the risk of a runaway should be considered. The term semi-batch refers to a system where one reactant and, if necessary, a catalyst is initially charged to a batch reactor. A second reactant is subsequently fed to the reactor under conditions such that an upset in reacting conditions can be detected and the flow of the reactant stopped, thus limiting the total amount of potential energy in the reactor. 

A transient, is a passing event which may upset the reactor operation but does not physically damage the primary cooling envelope. Table 6.1-1 lists PWR transient initiating events that ha c been used in PRA preparation. Typical frontline systems that mitigate LOCAs and transients for a PWR are presented in Table 6.1-2. The frontline systems must be supported by support systems interactions between both are presented in Table 6.1 -3 for ANO-1 (Arkansas Nuclear Unit 1). 

Transient An upset of the reactor power-to-flow raiin... 

In all tests, the temperature in the first- and second-stage reactors was kept within the necessary temperature limits of 288°-482°C. Because the carbon monoxide concentration was low in many of the tests, the second stage was not used to full capacity as is indicated by the temperature rise in runs 23, 24, and 27. The temperature profile shows the characteristic rise to a steady value. With the space velocities used (<5000 ft3/ft3 hr), the temperature profile is fully developed in the first stage within 30.0 in. of the top of the catalyst bed. A characteristic dip in temperature was observed over the first 8-10 in. of the catalyst bed in all runs. This temperature profile may indicate the presence of deactivated catalyst in this region, but, until the catalyst can be removed for examination, the cause of the temperature drop cannot be determined. There is no evidence that this low temperature zone is becoming progressively deeper. It is possible that an unrecorded brief upset in the purification system may have poisoned some of the top catalyst layers. 

Fig.2. HBr conversion during catalyst life testing in single full-scale reactor tube showing high conversion throughout the test. The brief time at lower conversion was due to a unit upset.

The general material balance of Section 1.1 contains an accumulation term that enables its use for unsteady-state reactors. This term is used to solve steady-state design problems by the method of false transients. We turn now to solving real transients. The great majority of chemical reactors are designed for steady-state operation. However, even steady-state reactors must occasionally start up and shut down. Also, an understanding of process dynamics is necessary to design the control systems needed to handle upsets and to enable operation at steady states that would otherwise be unstable. 

Choose a multiple steady-state case and try to upset the reactor by changing AO, F, TO or TJ interactively during a run. Only very small changes are required to cause the reactor to move to the other steady state. Plot as time and phase-plane graphs. 

There are several advantages of the use of HPLC for process monitoring. First, HPLC provides both qualitative and quantitative information about a process. At the research or pilot reactor stage of development, real time monitoring increases research efficiency and provides the data for process optimization. Second, because HPLC permits continuous real-time monitoring of reactors or other process components, process upsets that might go... 

Figure 10 Real-time plot of reactor component concentrations in catalytic hydrogenation step illustrating early detection of baseline upset by on-line micro-HPLC.

Although the flowsheet shown in Figure 13.7a is very attractive, it is not practical. This would require careful control of the stoichiometric ratio of decane to chlorine, taking into account both the requirements of the primary and byproduct reactions. Even if it were possible to balance out the reactants exactly, a small upset in process conditions would create an excess of either decane or chlorine and these would then appear as components in the reactor effluent. If these components appear in the reactor effluent of the flowsheet in Figure 13.7a, there are no separators to deal with their presence and no means of recycling unconverted raw materials. 

Hydrogen plant and gas turbine generator are each a partial load of the nuclear reactor during normal cogeneration operations. Transient or upset conditions in one system may... 

Figure 1-7 presents the causes of losses for the largest chemical accidents. By far the largest cause of loss in a chemical plant is due to mechanical failure. Failures of this type are usually due to a problem with maintenance. Pumps, valves, and control equipment will fail if not properly maintained. The second largest cause is operator error. For example, valves are not opened or closed in the proper sequence or reactants are not charged to a reactor in the correct order. Process upsets caused by, for example, power or cooling water failures account for 11 % of the losses. 

As in the case of a batch reactor for commercial operation, a CSTR is normally used for a liquid-phase reaction. In the laboratory, it may also be used for a gas-phase reaction for experimental measurements, particularly for a solid-catalyzed reaction, as in Figure 1.2. The operation is normally one of steady-state, except for startup, shutdown, and operational disturbances or upsets, in which cases unsteady-state operation has to be taken into account. 

Loss of instrumentation cause chemical reaction upset and explosion within the reactor occurred. 

Process upset from power interruption caused reactor explosion addition atmospheric release next day which resulted in explosion. 

Startup following maintenance produced process upset in reactor, water found in process materials, reflux line blocked resulted in overheating in reactor. 

---