AFD Operation
AFDK-SVU01C-EN
41
and inlet guide vane position to achieve the desired chiller
capacity and pressure coefficient. At the heart of the
control is a match model that describes the relationship
between control parameters and actuators. This model
has converted a complicated multi variable control
problem to a system of algebraic equations. The equations
cannot be solved directly, so a binomial search algorithm
is used iteratively to find a solution. A new solution is
found every 5 seconds.
Startup
The starting speed for the AFD will vary depending upon
the pressure ratio across the compressor. For most starts,
the pressure ratio will be small and the AFD will start at its
minimum speed. The speed will be adjusted every
5 seconds in response to changing pressure ratio and load
requirements.
On startup, shell pressures and temperatures may not
correspond to saturated conditions. To avoid potential
surge on start, the boundary pressure coefficient will be
reduced by 0.2 below the last running condition and after
40 minutes adjusts itself towards the last running
condition. This allows for the stabilization of pressures
and water loop conditions. After reaching this condition
the control will do a re-optimization.
Figure 22.
Startup surge boundary
0.0
0.6
0.7
0.8
0.9
1.0
1.1
1.2
10.0 20.0 30.0 40.0 50.0 60.0 70.0 80.0 90.0 100.0
Pressure Coefficient Trajectory Start to Full Load
Running surge boundary
Pressure Coefficient
Surge boundary for
start up
Boundary increases by 0.1
after 30 minutes
IGV % Capacity
Full load
Re-optimization
The AF Surge Boundary Offset Coefficient is a user
settable parameter to be used for adjusting the surge
boundary either higher or lower. In addition to being user
settable, the surge control algorithm will periodically
readjust this boundary. This re-optimization will occur
when any of three different criteria are met.
1. After startup stabilization the control will re-optimize
unless the surge is detected in that time period.
2. Every 30 minutes, the control will compare the current
IGV position with the IGV position at the end of the last
re-optimization time and, if greater than the user
adjustable sensitivity, will re-optimize.
3. When the re-optimization timer expires.
The control is re-optimized by increasing the AF Surge
Boundary Offset Coefficient every minute until surge
occurs. When surge occurs, the control will go into surge
recovery until the surge flag is removed and all of the re-
optimization timers will reset.
Figure 23.
Boundary re-optimization
Surge Recovery
When surge occurs, the pressures in the evaporator and
condenser shells can become erratic. Surge recovery is
needed to force conditions out of this unstable operating
point. This is accomplished by reducing the pressure
coefficient every 90 seconds of continuous surge. In
addition, when the surge flag is set, the compressor speed
command is increased by 1 Hz every 5 seconds until the
surge condition clears. When the surge flag is removed,
the speed command will relax back to the speed needed to
raise the pressure coefficient to the new surge boundary.
Surge Detection
Surge detection control logic monitors changes in
compressor motor current. A surge occurrence leaves a
characteristic motor current signature as shown in
. This signature is formed because the
transitory pressure breakdown between the condenser
and evaporator causes a sudden reduction in compressor
motor load. As the pressures equalize, the compressor
begins to quickly load, increasing the motor current.
0.0
0.6
0.7
0.8
0.9
1.0
1.1
1.2
10.0 20.0 30.0 40.0 50.0 60.0 70.0 80.0 90.0 100.0
Pressure Coefficient Optimization
Pres
sure Coefficient
Surge detected
IGV % Capacity
