Plant measurements

In addition to internal reactor measurements, plant measurements must include flows, pressures and temperatures around the reactors, and gas analyses. Flows, temperatures and pressures can be followed dynamically Ifom the eontrol system, whereas complete gas analyses are 

Before the validation of steady-state models, it must be realised that all experimental measurements have errors. 

In syngas plants only dry gas analyses are available and as a consequence, the concentration of water in the outlet gas from a steam reformer must either be determined by measuring the steam-to-dry gas ratio or it must be calculated using redundant measurements. Usually the last method is preferred due to large experimental uncertainties in the steam/dry gas measurement, where just 3% deviation in the measurement will correspond to about 10°C in temperature approach to the methane reforming reaction. 

Consistent values for the process variables aroimd the reactors must be known and for this purpose data reconciliation [72] [118] [351] is used to find a set of process variables, y, satisfying the basic mass and heat balances f not only around each individual process unit, but also 

In data reconciliation a mathematical method varies a set of independent variables x till the total sum of squares over all plant measurements, SQ, attains a minimum  

---

To determine if a process unit is at steady state, a program monitors key plant measurements (e.g., compositions, product rates, feed rates, and so on) and determines if the plant is steady enough to start the sequence. Only when all of the key measurements are within the allowable tolerances is the plant considered steady and the optimization sequence started. Tolerances for each measurement can be tuned separately. Measured data are then collec ted by the optimization computer. The optimization system runs a program to screen the measurements for unreasonable data (gross error detection). This validity checkiug automatically modifies tne model updating calculation to reflec t any bad data or when equipment is taken out of service. Data vahdation and reconciliation (on-line or off-line) is an extremely critical part of any optimization system. 

The third example shows how the uncertainties in plant measurements compromise the model parameter estimates. Minimal temperature differences, veiy low conversions, and hmited separations are all instances where errors in the measurements will have a greater impact on the parameter estimate. 

Plant-performance analysis reqmres the proper analysis of limited, uncertain plant measurements to develop a model of plant operations for troubleshooting, design, and control. 

Measurement Selection The identification of which measurements to make is an often overlooked aspect of plant-performance analysis. The end use of the data interpretation must be understood (i.e., the purpose for which the data, the parameters, or the resultant model will be used). For example, building a mathematical model of the process to explore other regions of operation is an end use. Another is to use the data to troubleshoot an operating problem. The level of data accuracy, the amount of data, and the sophistication of the interpretation depends upon the accuracy with which the result of the analysis needs to oe known. Daily measurements to a great extent and special plant measurements to a lesser extent are rarelv planned with the end use in mind. The result is typically too little data of too low accuracy or an inordinate amount with the resultant misuse in resources. 

FIG. 30-11 Plant measurements showing fluctuations around a mean value. 

FIG. 30-12 Plant measurements showing drift with time. 

FIG. 30-13 Plant measurements exhibiting step changes, drift, and random fluctuations. 

Assuming = 0 will potentially add bias to the interpretation of plant measurements. Further, the plant bias may to some extent mask the error in the measurements. While the designer may have envisioned a constant set of conditions or a specified time dependence, it is likely that the actual operation changes due to external factors. 

The following presents guidelines for identifying, validating, reconciling, rectifying, and interpreting plant measurements to remove some of the bias from the conclusions. 

Analysts should review the technical basis for uncertainties in the measurements. They should develop judgments for the uncertainties based on the plant experience and statistical interpretation of plant measurements. The most difficult aspect of establishing the measurement errors is estabhshing that the measurements are representative of what they purport to oe. Internal reactor CSTR conditions are rarely the same as the effluent flow. Thermocouples in catalyst beds may be representative of near-waU instead of bulk conditions. Heat leakage around thermowells results in lower than actual temperature measurements. 

Recommendations Plant measurements should be adjusted to close the constraints of the process. This adjustment shoiild be done on a component or subcomponent (e.g., atomic) basis. The adjustments should be done recognizing (at a minimum) the uncertainty in the measurements. While sophisticated routines have been developed for reconciliation, the vagaries of plant measurements may make them unsuitable in most applications. The routines are no substitute for accurate, precise measurements. They cannot compensate for the uncertainties and hmited information typically found in plant data. 

The skilled operator or the visiting service mechanic will have a working knowledge of the pressures and temperatures to be expected, but will not be able to make an accurate assessment of the actual conditions without plant measurements for comparison. The commissioning log (see Section 32.5) will show readings taken at that time, but only at one set of running conditions. 

FIG. 8 Potential difference between two Ag/AgCl electrodes in the stem of a soybean plant measured after adding to soil H2SO4. pH of soil was 3.0. Distance between Ag/AgCl electrodes was 5 cm. The plants were given water every other day and kept at 24°C. (From Ref. 3.)... 

For scale-up, the relation of the plant system (pit) to the laboratory or pilot plant measurement system (exp), in a simplified manner, becomes ... 

As discussed in Chapter 1, optimization of a large configuration of plant components can involve several levels of detail ranging from the most minute features of equipment design to the grand scale of international company operations. As an example of the size of the optimization problems solved in practice, Lowery et al. (1993) describe the optimization of a bisphenol-A plant via SQP involving 41,147 variables, 37,641 equations, 212 inequality constraints, and 289 plant measurements to identify the most profitable operating conditions. Perkins (1998) reviews the topic of plantwide optimization and its future. 

Live plant measurements will be fed to the model via the plant control computer. The model will then use the measurements and the target minimum gap to predict the alarm trigger point which will be communicated back to the control computer. This control computer is a conventional distributed control system (DCS), which has all the necessary software and displays for alarm handling and recording. The model itself will reside on a separate PC. Communications between the PC and the DCS will be subject to error checking and the system will default to the old fixed alarm value if a fault is detected. 

Analyzers for Near-line or At-line Plant Measurements... 

The design procedure used by Kosters, of Shell Oil Co., who developed this equipment, requires pilot plant measurements on the particular system of HTU and slip velocity as functions of power input. The procedure for scaleup is summarized in Table 14.5, and results of a typical design worked out by Kosters (in Lo et al., 1983, pp. 391-405) are summarized in Example 14.11. Scaleup by this method is said to be reliable in going from 64 mm dia to 4-4.5 m dia. The data of Figure 14.18 are used in this study. 

A.2 Analyzers for near-line or at-line plant measurements... 

Instruments which can monitor the important process variables during plant operation must be specified. These instruments must be capable of measuring the variables and should have an acceptable accuracy and repeatability of measurement, usually the latter attribute is more important than the former on chemical plant measurements. The instruments may be used for manual measurements or included in automatic control loops. Automatic alarms may also be required to indicate deviations outside acceptable limits. If possible, direct measurement of the process variable should be made, however it is often easier to measure a dependent variable, e.g. temperature measured as an indication of composition for distillation column top product. 

Neori, A., Clark, M., Beck, J., Huang, F., Kane, M.A. and Bitton, G. (1993) Heavy metal binding by natural waters and soluble exudates from micropropagated axenic wetland plants, measured with MetPLATE, a bioassay for rapid assessment of heavy metal toxicity, Amer. Soc. Limnol. Oceanog. (ASLO)-Soc. Wetland Scientists Ann. Joint Meeting, Univ. Alberta, Edmonton, May 31-June 3, 1993. 

---