Page 303 - Design of Simple and Robust Process Plants

8.2 Instrumentation 289
interpreted, as density changes caused by temperature or concentration deviations
are not recognized (except in commercial measurements where displacement meas-
urements are installed that are generally corrected for temperature and density).
Another problem is that, for correct installation, an installation length is required
upstream and downstream of the meter, and this is seldom the case. In specific
situations, temperature and density corrections are applied. Plugging of the lead
lines is another frequent cause for failure (see Table 8.1). The design errors in this
example are the selection of the instrument, the installation, and interpretation of
the signal.
A robust, reliable solution is found in the application of Coriolis meters for mass
flow; these are supplied in one embodiment with a density and temperature correc-
tions. These type of measurements also avoid installation of the lead lines.
Level measurements are typically performed by dP cells. The measurement is
effected by lead lines which are assumed to be filled with process liquid. During
start-up, these lines are often not filled, and during operation the density of the liq-
uid in the lines fluctuates with ambient temperature. Plugging of the lead lines is a
frequent cause of failure, especially when water is trapped in the lead lines, this
being a common cause of blocking during freezing weather. A more reliable solu-
tion is to use remote seals that avoid fouling and blockage, though the signal is then
affected by thermal expansion. Level measurements on dP cells are increasingly
replaced by capacity or ultrasonic measurements (as these are more reliable), or by
subtraction of two independent pressure measurements.
Pressure measurements are typically membranes, where pressure was transformed
into mechanical displacement that could be measured. In the past, this movement
was provided by a mechanical pointer, while in the next generation the displacement
was converted into a pneumatic signal available for pneumatic control. Clearly, it
was possible to convert this device into an electronic unit; this improved the accu-
racy because the mechanical displacement required was reduced, and so in turn was
the hysteresis of the membrane. Currently, the measurement of pressure is achieved
using a pizo crystal, which requires a minimum displacement and can resist
pulsed pressure; this is altogether a more robust design. Installation is carried out at
the vessel in order to avoid lead lines, the result being improved accuracy and much
greater reliability.
Temperature measurements also underwent evolution, starting with mechanical
displacement devices such as bimetallic or thermal expansion-based units. The ther-
mocouple, using different types of material combinations, was a good successor in
the electronic time-frame. For an adequate measurement, those materials were se-
lected with a relatively high emf / C, such as Fe-constantan. The cold junction of a
thermocouple suffered from ambient temperature variations. The resistance ther-
mometer partly replaced the thermocouple due to its greater accuracy, its disadvan-
tage being that a three-wired system is required. In the digital age with completely
sealed ICs being used for measurements, the thermocouple is again in focus. Cold
junction temperature variations are compensated with the local IC, while the emf
versus temperature curve is brought into the IC.