Pilot-Plant Experiments

Cycles Design methods for cycles rely on mathematical modeling (or empiricism) and often extensive pilot plant experiments. Many cycles can be easily analyzed using the methods described above apphed to the collection of steps. In some cycles, however, especially those operated with short cycle times or in shallow beds, transitions may not be very fully developed, even at a periodic state, and the complexity may be compounded by multiple sorbates. 

Pilot plant experiments represent an essential step in the investigation of a process toward formulating specifications for a commercial plant. A pilot plant uses the microkinetic data derived by laboratory tests and provides information about the macro kinetics of a process. Examples include the interaction of large conglomerates of molecules, macroscopic fluid elements, the effects of the macroscopic streams of materials and energy on the process, as well as the true residence time in the full-scale plant. 

Pilot plant experiments vary over a wide range, aeeounting for industrial eonstraints (e.g., duration of operation, eontrol parameters, equipment reliability, and impurities in the raw materials). Seale-up problems are investigated during pilot plant experiments. A pilot plant is an experimental rig, whieh displays the part of the operation that eorresponds to an industrial plant. It allows for simultaneous analysis of the physieal and ehemieal meehanisms. A pilot plant is indispensable for measuring the extent of the possible interaetions between these two types of meehanisms. It ean be small to minimize extraneous eosts sueh as the total operation eost as well as other eonstraints. 

Four pilot plant experiments were conducted at 300 psig and up to 475°C maximum temperature in a 3.07-in. i.d. adiabatic hot gas recycle methanation reactor. Two catalysts were used parallel plates coated with Raney nickel and precipitated nickel pellets. Pressure drop across the parallel plates was about 1/15 that across the bed of pellets. Fresh feed gas containing 75% H2 and 24% CO was fed at up to 3000/hr space velocity. CO concentrations in the product gas ranged from less than 0.1% to 4%. Best performance was achieved with the Raney-nickel-coated plates which yielded 32 mscf CHh/lb Raney nickel during 2307 hrs of operation. Carbon and iron deposition and nickel carbide formation were suspected causes of catalyst deactivation. 

A reaction is to be carried out in an agitated vessel. Pilot-plant experiments were performed under fully turbulent conditions in a tank 0.6 m in diameter, fitted with baffles and provided with a flat-bladed turbine. It was found that the satisfactory mixing was obtained at a rotor speed of 4 Hz, when the power consumption was 0.15 kW and the Reynolds number 160,000. What should be the rotor speed in order to retain the same mixing performance if the linear scale of the equipment is increased 6 times What will be the power consumption and the Reynolds number ... 

Two samples, one (Sample C) with Cr Oj, the other (Sample A) without any catalytic metal oxide, showed no release at all below 530°C. Judging from the pilot plant experience with FcjOj-containing catalysts, these two are not expected to be able to Action as SO transfer catalysts. 

Based upon laboratory and mini- or pilot-plant experiments process flow-sheets are prepared that also take environmental and safety aspects into consideration. Studies on these aspects are carried out in parallel with chemistry and engineering research in the form of so-called Integrated Process Development (see Heinzle and Hungerbiihler, 1997). 

A factory was designed and built specifically for the finasteride process. Performance through the dehydrogenation was comparable to our best pilot plant experience. Laboratory runs of the final step, however, revealed a problem the product was pink. There was obviously some difference in performance that we had not recognized. 

Cycles Design methods for cycles rely on mathematical modeling (or empiricism) and often extensive pilot plant experiments. Many... 

The type of agitator and tank geometry required to achieve a particular process result, is determined from pilot plant experiments. The desired process result may be the dispersion or emulsification of immiscible liquids, the completion of a chemical reaction, the suspension of solids in a liquid or any one of a number of other processes [Holland and Chapman (1966)]. 

Thus there are enough uncertainties in the kinetics in most chemical reaction processes that we almost always need to resort to a simplified model from which we can estimate performance. Then, from more refined data and pilot plant experiments, we begin to refine the design of the process to specify the details of the equipment needed. 

Known scale-up correlations thus may allow scale-up even when laboratory or pilot plant experience is minimal. The fundamental approach to process scaling involves mathematical modeling of the manufacturing process and experimental validation of the model at different scale-up ratios. In a paper on fluid dynamics in bubble column reactors, Lubbert and coworkers (54) noted ... 

In the first case, as discussed in Sec. VII below, a retrospective review of multiple batch records can provide considerable insight to support a defined PAR. A similar approach might involve a spreadsheet that summarizes critical parameter values for a series of R D lots when preparing to transfer the technology to R D s production colleagues. Often such retrospective data can be reinforced where gaps occur by some prospective laboratory or pilot plant experiments. 

Batch fermentations result in high product concentration (120-150 kg/m3) but in low productivity (2 kg m Iff1). Dramatic improvements in productivity (20-80 kg m Iff1 in laboratory- and pilot-plant experiments) were obtained by using cell recycle via MF or immobilized cells, at the expense of lactate concentration in the effluent (usually lower than 50 kg/m3). 

The design of RD is currently based on expensive and time-consuming sequences of laboratory and pilot-plant experiments, since there is no commercially available software adequately describing all relevant features of reactions (catalyst, kinetics, holdup) and distillation (VLE, thermodynamics, plate and packing behavior) as well as their combination in RD. There is also a need to improve catalysts and column internals for RD applications (1,51). Figures 8 and 9 show some examples of catalytic internals, applied for reactive distillation. 

The coke deactivation exponent n, is typically estimated from riser pilot plant experiments at varying catalyst contact time for different catalyst types. A value of n of 0.2 was found for REY catalyst data base. For USY and RE-USY catalysts n was estimated to be 0.4. 

This insight provided by the model and confirmed by the pilot plant experiments led to modifications in the large plant reactor. Liquid was returned to both ends and to the middle of the vessel. These simple plumbing modifications permitted the plant to use higher pressure setpoint ramp rates and at the same time reduced the frequency of disk ruptures. The higher TML yields and the shorter batch times increased production rates. 

Recently, one of the most practical results of these studies has been the ability to design pilot plant experiments (and, in many cases, plant-scale experiments) that can establish the sensitivity of process to macroscale mixing variables (as a function of power, pumping capacity, impeller diameter, impeller tip speeds, and macroscale shear rates) in contrast to microscale mixing variables (which are relative to power per unit volume, rms velocity fluctuations, and some estimation of the size of the microscale eddies). 

---