Page 3 - untitled

ctllp05044Mistakes.qxd 4/18/2005 11:29 AM Page 3

N ... AND HOW TO AVOID THEM 3

of the fluid, so they will sometimes comes to temperature instrumenta- products would eventually dwarf
work where magmeters fail. tion, for example, training costs any savings in equipment costs.
An even more challenging appli- and purchasing effort are reduced A fast sensor can also be worth
cation is the measurement of pH in when then vendor offers compati- the extra cost. If the process
a caustic liquid such as the slurries ble probes and transmitters togeth- requires a rapid succession of
found in paper mills. A general-pur- er as a package. heating and cooling cycles, the
pose pH probe made of corrodible temperature sensor must be able
materials might not only generate P Paayyiinngg ttoooo mmuucchh ((oorr ttoooo lliittttllee)) to generate a reading before it’s too
inaccurate data, it might die alto- Correct sensor selection is also a late to be of any use. Despite their
gether, sometimes within a matter matter of balancing cost against cheaper pedigree, thermocouples
of days. Some probes, such as performance. When there’s a tend to respond faster than RTDs—
those offered by ABB, are specifi- choice of equally effective tech- so if speed is the only important
cally designed for such tough envi- nologies, the right choice is gener- performance issue, choose a
ronments. They can double, triple, ally the cheapest one that gets the thermocouple.
and even quadruple probe life in job done.
many applications. Temperature instrumentation is a >> Mistake #2:
The trick is to find the right tech- classic example. The two dominant Installing sensors
nology for the application, or to technologies are resistance tem- incorrectly
choose instruments that span a perature detectors (RTDs) and
broader range of solutions. For thermocouples. An RTD consists of P Pllaacceemmeenntt
2
example, new digital technologies a metal plate or rod through which The best sensor can yield disap-
allow some flow meters to solve a current is passed. The resistance pointing results if not installed cor-
many more flow problems than that the current encounters varies rectly. Magmeters, for example,
their predecessors. with the temperature of the metal. A tend to generate noisy signals if the
Instrumentation vendors can be thermocouple consists of two dis- flow they’re measuring is turbulent.
of help in avoiding the technology similar metal wires joined together Bends, junctions, and valves in a
mismatch mistake. The best ven- at one end. The voltage between pipe can all cause turbulence, thus
dors train their sales people to the unjoined ends varies with the magmeters work best when
assist with sensor selection and temperature of the joint. Both yield installed in sections of straight
1 provide clients with easy-to-use voltages that can be electronically pipe.
selection guides. Some even offer interpreted to indicate the tempera- Temperature sensors are also
extensive look-up tables based on ture of the surroundings. sensitive to placement. Even a
product number, application, and Thermocouples are generally highly accurate RTD tucked in the
serial numbers of past installations— cheaper, though less accurate than corner of a mixing chamber will
an especially useful service when RTDs. If the application does not only be able to detect the tempera-
replacing older products. require particularly tight tempera- ture of its immediate vicinity. If the
Finding all the right parts can ture control, an inexpensive thermo- mixing of the material in the cham-
also be a challenge. Some instru- couple and a well-tuned PID loop ber is incomplete, that local tem-
ments require specific housings, should do the trick. But for process- perature may or may not represent
mounting hardware, and transmit- es that will only work correctly at the temperature of the material
ters to forward the sensor’s data to very specific temperatures, it would elsewhere in the chamber.
the controller. The right vendor can be a mistake not to pay for the Local temperature issues are
make all the difference by provid- greater accuracy that an RTD the classic mistake that home heat-
ing the entire assembly under a affords. The cost of scrapping a ing contractors often make when
single catalog number. When it batch of under-cooked or scorched installing household thermostats. A