BATCH PROCESS CONTROL
Batch processing is whereby a sequence of one or more steps, either in a single vessel or multiple vessels is performed in a defined order yielding a specified quantity of a finished product. Large production runs are achieved by repeating process steps on a predetermined schedule.
The key of using batch process as a method of manufacture is ensuring consistency of the product according to each specification while maximizing the available reactors and equipment. That is why most of the units in the batch process are organized such that that the plant is able to produce a range of products despite the limited amount of process equipment available.
One of the benefits of batch processing is that you have reduced inventories and shortened response times in order to makes specialty products as compared to continuous process.
Since it is not possible to blend multiple batches in order to obtain the desired quality, there are controls built into the process to ensure that the product quality are met during by each single batch and that the variances are minimized or within the acceptable limits set by Quality Control.
Detection of when a batch should be terminated (end point) may be performed by inferential measurements of product quality if direct measurements are not feasible. Tracking the end point of a batch process (which may be a function of time) is challenging since there is no steady-state operating point are also there are wide operation ranges due to frequent start-up and shutdown.
Setting the end point is determined by carrying out several lab scale models in Research & Development (R&D) in order to ensure that all variables are accounted for and limits are tested to see what happens is the process goes past the end point and tasks to remedy the situation.
Some of the challenges that might be observed while implementing control during a batch are:
1. Time varying characteristics. There is no steady-state operating point and transition in the controlled variable might be larger compared to typical excursion on continuous systems. Batch characteristics can change from run to run; even the process chemistry may change over a period of months due to changes in the product specifications.
2. Nonlinear behavior. Because of potentially wide range of operation, lineraized models may be inaccurate and inadequate for controller design. Batch chemical reaction rates may have a nonlinear dependence on temperature and concentration. Also there might be a nonlinear relationship between heat transferred from a reactor and the flow rate of the cooling medium.
3. Model inaccuracies. Often mechanistic or fundamental models are not available for batch processes thus limiting the ability to design and tune controllers.
Due to this, experiments are carried out to determine the actual model and based on this data, and then models are created. These models should be reviewed often in cases where the raw material change or equipment changes since reactors operate differently.
4. Sensors. For batch processes, on-line sensors are not available or are inaccurate due to the wide operating ranges during manufacturing. These infrequent samples are analyzed by the Plant laboratory. The inability to measure a process variable in real time reduces the safety margin for a process, leading to an undesirable operating condition.
However this can be redeemed by creating a model that plots on how far the reaction is proceeding as a function of time that has a built in end point that the plant operator can use.
5. Unmeasured disturbances. Operator error (opening the wrong valve, wrong tank feed chosen), fouling of vessel walls and heat transfer surfaces and raw material impurities are sources of major disturbances.
Since most of the disturbances are due to operator error, ensuring that the plant operators are adequately trained will lead to a major reduction of these errors. Also having a Standard Operation Procedure (SOP), that used sequential logic that encompass both normal and abnormal process operations would reduce disturbances caused by operator error.
6. Irreversible behavior. It is often hard to reverse the effects of product properties such as molecular weight distribution in a polymer or electrical properties in a semiconductor. Thus care should be taken during the batch process that the end point is always achieved that to ensure that the product meets is quality specifications.
Despite the challenges that face batch process control, a lot of industries use batch processes for production of their products thus giving them flexibility if they add more products in their portfolio in the future.
