POWER February 2011 - 32

INSTRUMENTATION & CONTROL
Morgantown's relatively low-cost generation
placed economic priority on environmental
and capacity goals. Unit maneuvering
speed affected economics briefly during the
spring and fall months and was not a high
priority. The controls were adjusted for operation
at 5 MW per minute.
Unit Ramp Rate Performance
Mechanical alterations took place on the pulverizers
mostly due to coal supply changes.
As a result, the DCS-mill modeling (originally
designed for the typical coals) became
more and more obsolete due to the changing
coal supply. The modeling estimated fuel
Btus released in the furnace vs. coal feeder
speed based on pulverizer motor amps and
was utilized to keep O2
, airflow, and feedwater
coordinated.
With increased variability of fuel supply,
the control of furnace parameters degraded.
One symptom was the increasing variability
of waterwall outlet temperature toward its
low and high alarm limits during load ramps.
Due to these limitations, the operations team
would frequently provide assistance by adjusting
the waterwall temperature setpoint
bias. In extreme cases, the operator would
place the firing trim controller in manual
mode and thereby assume waterwall and
main steam temperature control. Another
way to limit temperature variation was to
reduce the MW rate-of-change, which was
done with increasing frequency.
The final and most significant mill modification
was the addition of a variable-frequency
drive to power a new exhauster fan motor
while retaining the original motor to drive
the mill itself. The pulverizer upgrade also
included the addition of new bowls for the
dynamic classifier and new vane wheels. Before
the upgrade, the pulverizers were limited
in airflow capability. This resulted in a very
narrow range of coal feed during which the
exhauster inlet damper was modulated.
Full open damper position was reached at
a coal flow equating to mid-boiler load, and
the air/fuel ratio decreased as coal flow increased
due to the lack of exhauster fan capacity.
The mill upgrades provided increased
mill capacity as well as sufficient exhauster
fan capacity to control air/fuel ratio at the optimum
value throughout the entire coal feed
range. Mill fuel delivery response time was
reduced, eliminating the need for the previous
variable mill modeling techniques.
MW Ramp Rate Improvement
Project
Plant management desired improvement of
MW ramp rate for each steam unit because
physical changes to the equipment had
caused a degradation of MW ramping perfor32
mance.
Market conditions also converged to
increase the economic benefits of more rapid
load dispatch. Due to these factors, a ramp
rate improvement project was commissioned
to apply economical control changes for Unit
2. The primary technology employed for this
project was Invensys Connoisseur model predictive
control, an advanced process control
solution.
Ramp Rate Limitations Prior to
Control Modifications
Figure 1 presents 10 hours of load ramp
data compiled during December 2008 priwww.powermag.com
or
to the control and optimizer modifications
of the ramp rate project. An operator
increased the ramp rate limiter from 3 to
5 MW/min. This trend presents the load,
waterwall trim controller measurement,
setpoint and output, firing master output
and feedwater demand, north and south
superheat temperature and setpoint, four
desuperheat spray valve demands, and the
spray differential pressure (dP) controller
demand.
Load ramping caused large variations in
waterwall outlet temperature that approached
the high and low temperature alarm limits.
POWER | February 2011
1. Ramp rate limitations. Steam temperatures during dispatch at 5 MW/min are shown
(green arrow) prior to the installation of control modifications. While the north superheat temperature
was controlled close to the setpoint, the south superheat temperature frequently dipped to low values
(green arrow). Load ramping (red arrow) caused large variations in waterwall outlet temperature
(orange arrow) that approached the high and low temperature alarm limits. The white vertical time line
indicates a south superheat temperature of 962F. Courtesy: GenOn Energy
2. Temperature override protection. The desuperheater inlet steam temperature
override is shown during dispatch in a 1-hour portion of Figure 1. In this case, the override
controller responds to the highest of the desuperheater inlet temperatures (green arrow) and
reduces the waterwall outlet temperature controller output (violet arrow) if the desuperheater
inlet temperature exceeds the override setpoint. Courtesy: GenOn Energy

http://www.powermag.com

POWER February 2011

Table of Contents for the Digital Edition of POWER February 2011