POWER February 2011 - 48

INSTRUMENTATION & CONTROL
Thermocouple Response Time Study
for Steam Temperature Control
One important factor in power plant control system performance is the
response time of the process measurement used in the control system.
The response time of boiler steam temperature sensors and thermowells
is examined, as is those sensors' and thermowells' effect on
desuperheater temperature response time and, therefore, steam temperature
control performance.
By Cyrus Taft, Taft Engineering Inc.; John Sorge, Southern Company Generation; and Jackson Willis, Roy C. Martin & Co.
T
emperature measurements are used
extensively in power plants to monitor
and control many important
processes. Some of the most critical temperature
measurements are of steam temperatures
in the boiler, in particular the final
superheater outlet temperature and the
desuperheater outlet temperature.
In a typical drum boiler, the dry saturated
steam leaving the drum passes through
multiple superheating sections before it
leaves the boiler. A very common temperature
control method is to inject water
into the steam between superheater sections
to cool the steam. The steam and
water temperatures are commonly used in
a cascade steam temperature control system
in which the final temperature is the
measured variable for the outer control
loop and the desuperheater outlet temperature
is the measured variable for the
inner loop.
In any control loop, the response time
of the process is an important factor in the
performance of the loop. Faster-responding
processes are generally easier to control
tightly. Temperature measurements in
steam boilers and many other processes
have inherently slower response times than
many pressure and flow measurements.
The total response time of the temperature
measurement consists of the response
time of several components in the temperature
measurement system. When measuring
temperature in high-pressure steam
lines, the relatively fragile temperature
sensors are installed in thermowells to
protect the sensors from the harsh steam
conditions. The thermowells are heavy
metal tubes and, consequently, slow down
the temperature response at the sensor.
Other factors in the overall temperature
measurement response time include the heat
transfer from the steam to the thermowell,
48
www.powermag.com
conduction across the thermowell, heat transfer
from the thermowell to the sensor sheath,
and heat transfer to the sensor itself. If the
sensor is not fully inserted in the bottom of
the well, the heat transfer between the well
and the sensor will be significantly degraded.
Making Superheated Steam
In a drum boiler, steam is generated in
the furnace waterwall tubes, and the boiling
steam/water mixture is returned to the
drum. In the drum, the steam is separated
from the water before leaving the drum on
its way to the superheater. The superheater
is arranged in multiple sections with
the first section called the primary superheater.
There may be one or two additional
superheater sections, depending on
the particular boiler's design, as assumed
in this discussion. Between the two sections
of superheaters is a desuperheater in
which water from the feedwater system
is injected through spray nozzles into the
steam for steam temperature control.
As the water evaporates, it cools the
steam and provides a means of temperature
control. Evaporation happens very
quickly, provided the spray nozzles atomize
the spray water well. The temperature
just downstream of the desuperheater after
evaporation is completed is measured
and referred to as the desuperheater outlet
temperature.
After the steam leaves the desuperheat1.
Great expectations. This diagram shows the expected time response of desuperheater
components. Courtesy: Southern Company Generation
Valve demand
Thermowell temperature
Valve position
Measured temperature
Actual steam temperature
0.8
0.6
0.4
0.2
-0.2
-25
25
50
Time (seconds)
POWER | February 2011
75
100
125
Response

http://www.powermag.com

POWER February 2011

Table of Contents for the Digital Edition of POWER February 2011