Manual de Aplicacion Completo
Manual de Aplicacion Completo
Manual de Aplicacion Completo
English
Original Instructions 06-2018 Enter document number (Issue
1)
Table of Contents
1. INTRODUCTION ...................................................................................................................... 1
1.1 Scope ................................................................................................................................ 1
1.2 Application Manuals .......................................................................................................... 1
1.3 Safety ................................................................................................................................ 2
5.1 Description....................................................................................................................... 33
5.2 Transition Types .............................................................................................................. 34
5.3 PTC Timers ..................................................................................................................... 35
5.3.1 Programmed Transition Time Delay (TDPT) ....................................................... 35
5.3.2 Normal to Emergency Transfer Time Delay (TDNE) ........................................... 35
5.3.3 Emergency to Normal Retransfer Time Delay (TDEN) ....................................... 35
5.3.4 Maximum Parallel Time (TDMP) .......................................................................... 35
5.4 PTC Functions................................................................................................................. 36
5.4.1 PTC Mode selection ............................................................................................. 36
5.4.2 Test with Load ...................................................................................................... 36
5.4.3 Extended Parallel Enable ..................................................................................... 36
5.4.4 Commit to Transfer............................................................................................... 36
5.4.5 Utility Breaker Open Point .................................................................................... 37
5.4.6 Fail to Disconnect Enable..................................................................................... 37
5.4.7 Fail to Sync Open Transition Retransfer .............................................................. 37
5.4.8 Source Availability ................................................................................................ 37
5.4.9 Exercise ................................................................................................................ 40
APPENDIX B. PI TUNING........................................................................................................... 67
1.3 Safety
Safety should be a primary concern of the facility design engineer. Safety involves two aspects: safe
operation of the generator set itself (and its accessories) and reliable operation of the system. Reliable
operation of the system is related to safety because equipment affecting life and health is often
dependent on the generator set, such as hospital life-support systems, emergency egress lighting,
building ventilators, elevators, fire pumps, security and communications.
Refer to any relevant codes and standards for information on applicable electrical and fire codes
around the world. Standards, and the codes that reference them, are periodically updated, requiring
continual review. Compliance with all applicable codes is the responsibility of the facility design
engineer. For example, some areas may require a certificate-of-need, zoning permit, building permit or
other site-specific certificate. Be sure to check with all local governmental authorities early in the
planning process.
NOTICE
While the information in this and related manuals is intended to be accurate and useful, there
is no substitute for the judgment of a skilled, experienced facility design professional. Each
end user must determine whether the selected generator set and emergency/standby system
is proper for the application.
• When the slip frequency method is used the control does not try to drive the generator set
waveform to be in sync with the utility but drives the frequency to be offset from the utility
frequency by some “slip” frequency, typically on the order of 0.1 Hz so that the utility and
generator set will periodically be in sync with each other. (With a slip frequency of 0.1 Hz the
phase difference will complete one cycle every 10 seconds.)
When the slip frequency method is used the control calculates the optimum time to close the breaker
based on the breaker closing time and the rate of change of phase angle to close at a phase
difference of 0 degrees. Note that a higher slip frequency results in faster synchronizing but can result
in high current spikes at the instant of closing a breaker. If done repeatedly this can lead to damage to
the alternator. This tradeoff of fast synchronizing vs minimizing stress on the alternator needs to be
considered. In a standby application where speed of synchronizing is of paramount importance and
there will be few synchronizing events over the life of the generator set it is permissible to have a
higher slip frequency. In applications where the generators will synchronize and close to a live bus
often over its life it is important to minimize slip frequency (typically to values of less than 0.1 Hz) in
order to minimize stress on the alternator. Consult the generator set manufacturer for more insight into
this.
3.1.4 Setup
The InPower service tool can be used to set up the speed control method to be either phase match or
slip as discussed above. The control can also be set up so that an external synchronizer can manage
this function using the speed bias input. Similarly the synchronizer voltage control method can be
configured in InPower for either a voltage match function or to allow an external device to manage the
voltage matching using the voltage bias input. Figure 3 shows the InPower screen for setting the
synchronizer speed and voltage control methods as well as the slip frequency settings.
3.2.1 Setup
Figure 5 displays the InPower setting for the sync check function. Note that there is a setting for a
predictive sync check however that function is currently not enabled.
3.3.3 Backup
In case the arbitration network is compromised, each genset has a First Start Backup time
programmed to enable generator sets to close to the bus if the First Start Arbitration process fails. If
the First Start Backup Timer expires and the bus is still not live the genset will close to the bus. The
First Start Backup Timer can be set with InPower within a range of 3 to 120 seconds (see Figure 6 on
page 13). To prevent the possibility of multiple generator sets simultaneously closing to the dead bus
it is required that each generator set have a different first start backup time programmed. The normal
process typically takes less than 1.5 seconds so a backup time of 5 seconds for the first generator set
will be sufficient, with a difference of 5 seconds between each subsequent generator set’s backup
time.
FIGURE 6. INPOWER SETTINGS FOR ARBITRATION SPEED THRESHOLD, FIRST START BACKUP TIME
AND LOSS OF VOLTAGE SENSING ENABLE
3.3.6 Setup
First Start Arbitration Threshold and First Start Backup time can be configured in InPower. The Loss of
Voltage Sensing function can be enabled or disabled using InPower. Figure 7 on page 13 displays
these settings.
First Start Arbitration Threshold can be set from 50% to 90% of engine speed, with a default of 90%.
Setting the Arbitration Threshold to a lower value causes arbitration to start earlier which reduces the
time it takes for the first genset to close to the bus. This does create a slight risk that a genset will win
arbitration and then fail to come up to speed which would result in a longer first start time. Note that no
generator set will close to the bus before the engine speed is 90% of nominal regardless of what the
arbitration threshold is.
First Start Backup time is used to enable generator sets to close to the bus if the First Start Arbitration
process fails. If the First Start Backup Timer expires and the bus is still not live the genset will close to
the bus. The First Start Backup Timer can be set with InPower within a range of 3 to 120 seconds.
Each generator set should have a different time programmed to prevent the possibility of multiple
generator sets simultaneously closing to the bus.
The Loss of Voltage Sense function can be enabled or disabled in InPower. It is enabled by default. It
should be disabled when the analog load sharing lines are not used, such as when load sharing is
done by droop or through an ILSI of through the bias inputs from a 3rd party control.
“Fail To Close” and “Fail to Open” faults are provided if the breakers fail to respond to these
commands. PTC Applications also provide a Fail To Disconnect fault which will indicate that neither
source will open when commanded to do so. This fault could be mapped to a configurable output
which could be used to cause external action.
There are also configurable inputs available to indicate that a breaker has tripped. Breakers will also
often have a racked out contact. There is not a defined input on the control for this but a customer
configurable input may be used.
An additional diagnostic “Genset and Bus Voltages Out of Calibration” is activated when the sensed
alternator and bus voltages don’t agree within a value defined by the “Gen to Bus Voltage Cal Check”
trim and the breaker(s) are closed. This applies to whichever bus the generator closes to, whether it is
a utility or a generator paralleling bus.
3.4.3 Setup
Figure 8 and Figure 9 display InPower settings for breaker control and diagnostic functions. Note that
the functionality is described above.
In a PID control a reference is generated based on a desired set point and compared with the
measured value. The difference between the two is the error. The error will be a positive number if the
setpoint is higher than the measured value. It will be a negative number if the setpoint is lower than
the measured value. The error, after being scaled by proportional, integral and derivative functions,
determines the output of the control loop. The control continuously makes adjustments to bring the
error to zero, meaning that the measured value is equal to the setpoint.
For example an engine with a setpoint of 1800 RPM that is operating at 1790 RPM will have an error
of 10 RPM. If that same engine were operating at 1820 RPM it would have an error of -20 RPM.
In a PID control the integral is the sum of error terms over the time the control has been running.
When an engine first starts the error will be a positive number as the speed is low so the integral
increases. As the speed overshoots the setpoint, the error becomes negative so the integral will
decrease. As the engine settles to 1800 RPM the error becomes zero. At steady state the proportional
and derivative terms also become zero. The control output is determined only by the integral term.
The integral will be at the value corresponding to the output torque necessary to maintain 1800 RPM
at whatever load is on the engine.
As loads cycle on and off of the generator set the speed will fluctuate creating positive and negative
error terms. The error terms will add to or subtract from the integral so that the torque will be sufficient
to carry the connected load and maintain a speed of 1800 RPM.
When a generator set is paralleled to another source (either another generator set or a utility
transformer) the governor is not able to control speed because the rotor is magnetically coupled
(locked) to the other source. By controlling fuel however, the governor is able to control the torque of
the engine which directly controls the real power (kW) output of the paralleled generator set.
The “offset” shown in the block diagram (Figure 10) is used as an additional input to the governor
control loop to adjust the real power output of the engine to the desired kW value. The offset adjusts
the error term adding to or subtracting from the governor’s integral updating engine torque to produce
the desired kW output.
The governor setpoint is the rated value plus the offset. This rated value keeps the engine running at
or close to nominal frequency while the offset is used for maintaining torque to produce the correct
real power output.
The voltage regulator control loop is analogous to the governor control loop. The basic function of the
regulator is to adjust excitation current to maintain genset output voltage, but what the excitation
actually controls is magnetic flux. It takes a certain amount of flux to maintain 480 Volts at no load. It
takes more flux to maintain nominal voltage when the alternator is loaded. When paralleled to another
source the voltage regulator is not able to control voltage, however by controlling excitation current the
control is able to control the magnetic flux through the alternator. At constant voltage reactive power
output (VARs) is a function of flux, so in a paralleling application the voltage regulator, by controlling
the magnetic flux, controls the reactive power output of the alternator.
See appendix 2 for a discussion of PI tuning techniques.
As a reminder, the four basic paralleling algorithms are:
• Synchronizing – Adjusting the frequency, phase angle and voltage of the generator set’s output
waveform to match the waveform of another source (either another generator set or a utility
transformer) prior to closing the generator set breaker to parallel to another source.
• Load Sharing – generator sets paralleled (paralleling breaker is closed to the bus) with each
other in an isolated bus configuration (no utility) carrying equal parts of the load (kW and kVAR)
as a percentage of the generator set ratings while maintaining rated frequency and voltage on
the generator sets.
• Load Governing – Controlling how much load (kW and kVAR) the generator set carries when
paralleled with a utility transformer
• Droop – Varying frequency and voltage offset in a negative slope (ie increasing kW results in
decreasing speed offset) as a function of real and reactive power carried by a generator set
(respectively). The droop function can be used in standalone, isolated bus or utility paralleling
applications.
The offset is determined by what paralleling mode the generator set is in. Figure 11illustrates how the
generator set control chooses a frequency offset based on the speed control mode. Note that in this
case the term “isochronous” refers to running standalone at a constant speed, in which case the offset
is 0. The control uses similar logic in selecting the voltage control offset.
Load share is activated when the generator set paralleling breaker is closed and the utility breaker is
open. Note that when the control is configured as “isolated bus only” there is no utility breaker and
load share will be activated whenever the paralleling breaker is closed.
The load share function includes a ramp only when in isochronous load sharing, the generator set is in
automatic mode and the generator starts or stops based on the load demand function. When the
generator starts or stops manually there is no ramp in the load share function. Load demand starts
and stops will be discussed in a later function.
If the generator set is overloaded (frequency has deviated by more than 1 Hz) or reversed powered
the ramp is aborted and the control will share load at steady state.
4.1.3 Setup
There are several parameters that affect the load share function as displayed in Figure 13.
• Load Share kW Gain and Load Share kVAR Gain. These are the gains of the load share control
loop. In most applications it is appropriate to leave these gains at the default value of 1.0
however these gains may be reduced if there is instability in how the generator sets are sharing
load.
• Load Share kW Balance and Load Share kVAR Balance. These trims are used to adjust the
balance of load between generator sets. If for some reason it is desirable for one generator to
carry more than its share of the load or if the generator sets are not sharing load equally these
parameters can shift the balance of load between generator sets. The default setting is 0.
Increasing the value will cause the generator set to increase its share of the load. Decreasing
the value (making it negative) will cause the generator to decrease its share of the load.
• Load Share Ramp Load Time and Load Share Ramp Unload Time. This is the time will take for
a generator set to ramp load on or off when it connects to or disconnects from a generator
paralleling bus. This is used when a generator is started or shut down due to the Load Demand
function. The ramp is defined as the time will take for a generator set to ramp from 0% to 100%
or from 100% to 0% of its rated load. If it is only ramping up to 50% of its rated load the ramp
will take half the time. (For example if the trim is set to 30 seconds and the load the generator is
taking on is 50% of its rating the ramp will take 15 seconds.)
• Load Share Ramp kW Unload Level. When a generator set is unloading to shut down this is the
level at which the control will consider the generator set as unloaded and will open the breaker
during a ramp down.
While running in load share it is not permissible to modify the voltage and frequency adjust
parameters.
• Utility kW - This is often called “peak shave”. In this mode the control adjusts genset kW output
in order to maintain a fixed setpoint kW level on the utility. This is only applicable when the
paralleling application is either utility single or Power Transfer Control.
There are four modes of kVAR load governing.
• Genset kVAR - In this mode, the control regulates generator set kVAR output to a fixed setpoint
value. Genset kVAR can be set to a negative value, which represents operating with a leading
PF or absorbing VARs when using an internal setting (HMI, InPower or Modbus). When using an
analog input to set kVARs only settings from 0 to 100% (corresponding to lagging PF) are
allowed.
• Utility kVAR - In this mode, the control adjusts generator set kVAR output in order to maintain a
fixed setpoint kVAR level on the utility. This is only applicable when the paralleling application is
either utility single or Power Transfer Control. For utility kVARs only lagging PF operation is
allowed.
• Genset Power Factor - In this mode, the control adjusts generator set kVAR output to maintain a
fixed generator set power factor. Genset PF can be set to a negative value, which represents
operating with a leading PF or absorbing VARs, when using an internal setting (HMI, InPower or
Modbus). When using an analog input to set PF only settings from 0 to 1.0 (corresponding to
lagging PF) are allowed.
• Utility Power Factor - In this mode, the control adjusts generator set kVAR output to maintain a
fixed utility power factor. This is only applicable when the paralleling application is either utility
single or Power Transfer Control. For utility PF control, only lagging operation is allowed.
Consult the factory if operation at leading Power Factor is desired.
All load govern modes will maintain the generator set output within the generator set standby rating.
Maximum kW output will be 100% of kW standby rating unless limited by a maximum kW limit trim.
Maximum kW limits may be further restricted by derates. See the derate section for further discussion.
4.2.2 Setup
Figure 14 and Figure 15 show the InPower setup screens for the load governing function. Figure 14
shows the selection of load governing or droop for utility paralleling speed and voltage control, the
load govern kW and kVAR setpoint sources and load govern methods. The kW and kVAR setpoints
can also be set on this screen.
Figure 15 also displays the adjustment for proportional and integral gains for both load govern kW
and kVAR. Generally speaking these gains should be left at their default values. If there is instability in
load governing, that is if the kW or kVAR that the generator is putting out is oscillating, reducing the
proportional gain is something to try. Reducing the gain too much however, will result in a system that
is very slow to respond to changes in setpoint. Reducing the integral gain can also reduce oscillations
as well. Too high of an integral gain can result in overshooting a setpoint however too low of an
integral gain can result in a system that takes a long time to settle in to its setpoint. Refer to appendix
2 for a discussion of PI tuning.
4.2.3 Derate
Several of the PowerCommand Controls events and faults may be configured as a derate rather than
a warning or a shutdown. Each derate event will have some percentage by which to derate the
generator set while that event is active. For example, a derate of 20% would cause the generator set
to operate as if the standby kW rating were 80% of its original value.
When a derate becomes active while the generator set is operating in load govern mode, the derated
kW effectively becomes the new maximum standby rating. If the kW set point exceeds the derated
maximum kW rating the output will be limited to the derated maximum kW.
When a derate becomes active while the generator set is not operating in load govern mode then the
load dump function becomes active when the actual kW exceeds the derated maximum kW rating.
Derate applies only to kW. There is no derate for kVAR.
4.3.1 Setup
Droop speed and voltage control can be initiated by either an InPower setting or by a configurable
input. Figure 19 shows the InPower settings for the droop function. Notice that there are separate
settings for Utility Parallel speed and voltage control methods, as well as Isolated Bus speed and
voltage control methods. The configurable input takes precedence over the InPower setting, no matter
what paralleling mode the generator set is in. For example if a configurable input is configured for
voltage droop and that input is activated the generator set control will go into droop voltage control
mode even if constant voltage mode was selected in InPower.
FIGURE 21. FREQUENCY DROOP CHARACTERISTIC WITH FREQUENCY ADJUST SET TO A NEGATIVE
NUMBER
Voltage droop settings are a bit different as the control will allow the generator set to absorb VARs.
Figure 22 shows the voltage droop characteristic with the voltage adjust set to 0. The slope ‘A’
represents the voltage change corresponding to a change in kVAR from 0 to rated kVAR. Note that
rated kVAR is defined as 60% of the generators sets rated standby kVA. For example for a genset
rated at 1000 kVA, 0.8 PF, rated kVAR is 600 kVAR, or 60% kVAR per standby kVA.
In a utility paralleled application (where generator sets control output but not frequency or voltage)
lower percentages can result in instability because a slight change in frequency or voltage can result
is a dramatic change in output. In this case higher percentages can result in generator sets not
changing their kW and kVAR output as fast as they should, which is the inverse of the isolated bus
case.
With a soft load transfer, the control ramps load between the generator set and the utility based on the
Load Govern kW Ramp Load Time and Load Govern kW Ramp Unload Time. When the control
determines that the source that is to be disconnected has been unloaded it opens that breaker. When
disconnecting from the generator set the control will determine that the generator set has been
unloaded based on the kW unload ramp time. (Note that the breaker actually opens when the kW
target reaches 0. Refer to the load govern section for discussion of ramping loads from the gen to the
utility.) When disconnecting from the utility breaker the control will consider the utility unloaded based
on the Utility Unloaded Level set in InPower, shown in Figure 23 on page 34.
5.4.8.1 Sensors
5.4.8.1.1 Voltage and Frequency Sensors
The PowerCommand control has voltage and frequency sensors on both the generator set and the
utility. The PTC sensors are used to determine whether or not a source is acceptable. Source
acceptability is used to drive whether or not to start generator sets, whether or not to switch the loads
to a different source, and whether or not two sources can be paralleled. The sensors have some
hysteresis built in so that source doesn’t oscillate between available and unavailable. Figure 26 shows
the Pickup and Dropout Thresholds for the Undervoltage Sensor. When the voltage is above the
Pickup Voltage the source is considered available. If the voltage falls below the dropout voltage the
source is considered unavailable and won’t be considered available again until the voltage rises above
the pickup voltage. Pickup and dropout percentages can be set in InPower as shown in Figure 24 on
page 38 and Figure 25 on page 39
FIGURE 26. PICKUP AND DROPOUT THRESHOLDS FOR THE UNDERVOLTAGE SENSOR
5.4.9 Exercise
The exercise function is basically a timed test. An exercise can be initiated either by a configurable
input, a Modbus command or the scheduler. If initiated by a configurable input or Modbus command
the exercise time is set in InPower and the function uses the same “Test with Load” setting as the
Test function to determine whether or not a transfer of loads to the generator set should be included in
the exercise. Both the generator set and the PTC function must be in Auto mode for the Exercise
function to operate.
The scheduler allows an operator to select up to 12 repeating exercise programs, each of which can
be configured for “With Load”, “Without Load” or “Extended Parallel”. For each program the operator
can select the day of the week, the time to start, the duration of the exercise and whether or not the
exercise will be with or without load. The operator can also select the repeating frequency,(ie repeat
every week, every two weeks and so on up to every 5 weeks). The operator can also choose to have
the exercise repeat once a month on a specific day of the week, for example the operator can choose
the 2nd Wednesday of the month for an exercise program.
The scheduler uses the Real Time Clock. When setting up a schedule make sure the Real Time Clock
is correct and enabled. There is a diagnostic (fault code 1689) that indicates that power to the control
has been lost and the Real Time Clock must be re-set.
The operator is also able to configure exceptions to the scheduler. For example if one of the
programmed exercises falls on a holiday so it is not desired to have an exercise on that date the
operator can enter that date into the exception table and that will disable the exercise for that date.
If the exercise is interrupted it will not continue its run. For example if the utility fails while running on
the generator set the exercise will end and normal operation will resume so that PowerCommand will
re-transfer loads to the utility as soon as it becomes available. If the utility is not available at the time
of a scheduled exercise the exercise will be cancelled, even if the utility becomes available during
what would have been the exercise period.
Figure 27 displays the InPower scheduler. Methods to confirm that the exercise occurred include
monitoring engine run hours and number of starts.
2. After all of the generator sets have started and paralleled and an initial time delay has expired
the system calculates how much load is being supplied by the generator sets. If there is
excessive capacity on line, the system will shut generator sets down so that running capacity is
better matched to the load.
3. The system will continue to monitor capacity and load, starting and stopping generator sets as
configured. The kW thresholds at which generators are started and stopped and the order in
which generators are stopped and started depend on the MLD settings described below.
The Load Demand System settings, which will apply to all generator sets in the system, are displayed
on the right side of Figure 28. Individual generator set settings are displayed on the left side. The
general process for setting up the MLD systems is to first set the individual settings for each generator
set. After that is complete, the system settings can be written to one generator set, then by pressing
the “Synchronize System Settings” button at the bottom right of the screen, the system setting are
communicated to all of the generator sets on the S-CAN network.
When running in run hour equalization type the generator set priorities may change as the run hours
of a running generator set may surpass the run hours of a generator set that has shut down. When the
run hours of the running generator set surpasses the run hours of a stopped generator set by the “Run
Hour Differential” setting (shown on the right side of Figure 28) the MLD system will switch the priority
of the two generator sets. The stopped generator set will be started and run in parallel with the
running generator sets and the running generator set which has now been assigned a lower priority
will be load demand stopped.
The operator has the option to base the run hour equalization on a value that is different from engine
run time. That number can be entered in the “LD Genset Run Hour” setting shown on Figure 28.
Examples of why an operator may want to use a different run hour setting are a control board has
been replaced and doesn’t match actual generator run hours, or an operator may want to temporarily
change the priority of a generator set.
Figure 29 shows the InPower screen for the Load Demand Fixed Priority. The operator enters the
information for the fixed priority sequence on the left side of the screen. On the right side of the screen
InPower displays the priority sequence of the system. If the Load Demand Type is Fixed Priority then
the right side of the screen will match the left side. If Run Hour Equalization is used, the right side of
the screen will display the priorty setting assigned by the MLD system based on current run hours.
As with other MLD system settings, the priorty setting should synchronized after the individual
generator set settings have been written. Synchronization includes both writing and saving to other
generator sets. The priority setting only need to be written to one generator set and the will be
commmunicated to all of the other generator sets via the S-CAN network. MLD parameters may be
synchronized while generator sets are stopped or running.
When the relative method is selected the start and stop thresholds are based on the ratio of load to
capacity. The start threshold is set to the maximum ratio of load to online capacity. The stop threshold
is set to the minimum ratio of load to online capacity.
As an example, consider a system of 5 1 MW generator sets. In this example the Start Threshold is
set to 80% and the Stop Threshold is set to 60%. Figure 30 shows a histogram of loads and online
generator capacity during a sequence following a utility failure.
FIGURE 30. EXAMPLE OF MLD FUNCTION USING THE RELATIVE THRESHOLD METHOD
1. After a utility failure, all five 1MW generator sets start, so online capacity is 5 MW.
2. The system calculates that there is 100 kW of load. The stop threshold is set to 60%. This
means that the system will shut down generator sets as long as doing so will not cause the ratio
of load to online capacity to be above 60%. In this example the system will shut down four of the
generator sets, leaving an online capacity of 1000 kW. The load is less than 60% of the online
capacity (600 kW in this case), which causes the system to leave only one generator set running.
At this point the Next Start Threshold is 800 kW, 80% of online capacity. If the load increases
above 800 kW a second generator set will start.
3. The load increases to 900 kW. That exceeds the Start Threshold of 80% of online capacity, so a
second generator set starts. Now the online capacity is 2000 kW. If the load exceeds 80% of
capacity (1600 kW) an additional generator set will start. The kW level at which one of the
generators will shut down is 60% of what the capacity will be after the generator set shuts down.
If one generator shuts down the online capacity will be 1000 kW so a generator will shut down if
the load drops below 600 kW (60% of online capacity).
4. The load is now 2100 kW. Four generator sets are online so online capacity is 4000 kW. The
load to capacity ratio is 53%. Although this is below the 60% Stop Threshold a generator set
doesn’t shut down because the stop threshold is calculated based on what the online capacity
will be after a generator set shuts down. If a generator set shuts down the online capacity will be
3000 kW. The Stop Threshold is 60% of that or 1800 kW. When the load drops to 1800 kW, the
generator set shuts down.
The other method for determining the Start and Stop Thresholds is the Absolute Method. When the
absolute method is used the thresholds are based on surplus capacity, also known as spinning
reserve. The Load Demand Start Threshold is the minimum surplus capacity before starting the next
generator set. The Load Demand Stop Threshold is the maximum surplus capacity before shutting
down a generator set.
As an example, consider a system of 5 1 MW generator sets. In this example the Start Threshold is
set to 1500 kW and the Stop Threshold is set to 1800 kW. This means that when possible the MLD
system will maintain a surplus capacity in the range of 1500 kW – 1800 kW. Figure 31 shows a
histogram of loads and online generator capacity during a sequence following a utility failure.
FIGURE 31. EXAMPLE OF MLD FUNCTION USING THE ABSOLUTE THRESHOLD METHOD
1. After a utility failure, all five 1MW generator sets start, so online capacity is 5 MW.
2. The system calculates that there is 100 kW of load. The stop threshold is set to 1800 kW. This
means that the system will shut down generator sets as long as doing so will not cause the
surplus capacity to drop below 1800 kW. In this example the system will shut down three of the
generator sets, leaving an online capacity of 2000 kW. The surplus capacity is 1900 kW (2000
kW capacity – 100 kW load) which causes the system to leave two generator sets running
because shutting another generator set down would cause the surplus capacity to be less than
1800 kW. At this point the Next Start Threshold is 500 kW of load, which would result in 1500 kW
of surplus capacity. If the load increases above 500 kW a third generator set will start.
3. The load increases to 600 kW. That causes the surplus capacity to drop to 1400 kW, which is
below the Start Threshold of 1500 kW, so a third generator set starts. Now the online capacity is
3000 kW and the surplus capacity is 2400 kW. If the load exceeds 1500 kW this will cause the
surplus capacity to drop below 1500 kW and an additional generator set will start. If the load
drops below 200 kW a generator set will shut down as this will leave the system with a surplus
capacity in excess of 1800 kW after that next lower priority generator set has shut down.
4. The load is now 2400 kW. All five generator sets are online so online capacity is 5000 kW and
the surplus capacity is 2600. Although this is above the Stop Threshold of 1800 kW a generator
set doesn’t shut down because the stop threshold is calculated based on what the capacity will
be after a generator set shuts down. If a generator set shuts down the online capacity will be
1400 kW. When the load drops below 2200 kW, the generator set shuts down.
The basic sequence of operations for a Spare Capacity Request event is as follows:
1. Spare Capacity Request input is activated.
2. Start and Stop thresholds are adjusted as described above
3. Load Demand Start Delay initiates
4. After load demand start delay expires, next priority generator set(s) starts, synchronizes and
closes to the paralleling bus.
5. Spare Capacity available output becomes active
6. Load Demand Start delay for next priority generator set(s) starts
7. Operator or system adds load
8. After load is added, operator or system de-activates spare capacity request before load demand
start delay expires. (If load demand start delay expires before spare capacity request is removed,
an additional spare capacity request will be added.)
If the capacity requested exceeds the kW rating of a single generator set, two (or more if required)
generator sets will start simultaneously after load demand start delay expires, as described above.
7.1 Biasing
PowerCommand controls have analog inputs and outputs that can be configured as bias signals. The
3.X controls have both inputs and outputs on the base board. 1.X and 2.X controls have only bias
inputs on the Aux101.
In general, a paralleling module drives a bias output into a generator set control bias input to offset the
governor and voltage regulator setpoints so that the paralleling module can manage paralleling
functions. Bias signals are often used when paralleling generator sets from different manufacturers or
when the paralleling system controller is provided by a different company than the generator set
controller.
The PowerCommand bias inputs can be used to allow a 3rd party controller to control the generator set
paralleling functions. The PowerCommand 3.3 bias outputs can be used to control the paralleling
functions of a 3rd party generator set.
In applications that include several Cummins generator sets with one 3rd party generator set, a
PowerCommand 3.3 control can be used as a paralleling module with that 3rd party generator set to
enable paralleling of that generator set with the Cummins generator sets. Conversely, in an application
with several 3rd party generator sets and one Cummins generator set, a 3rd party paralleling control can
use the bias inputs on a 3.3 control to enable paralleling between the Cummins generator set and the
3rd party generator sets.
When PowerCommand controls with 3rd party equipment, the PowerCommand 3.3 control acts as
either the paralleling module or the generator set control for both the governor and the AVR. It is not
permissible for the PowerCommand control to act as the paralleling module for the governor and the
generator set control for the AVR, for example. “Hybrid” controls, in which one control acts as the
paralleling module for the governor and the generator set control for the AVR, are not allowed. The
PowerCommand 3.3 control needs to operate as either the paralleling module or generator set control
for both the governor and the AVR.
Bias range setup:
The range of the bias signal should be wide enough to cover the limits of the synchronizer, load share
and/or load govern functions. Limiting the range of the bias signal limits the effect that the bias signal
can have on the governor and/or AVR. Setting the limits too narrow may limit performance of load
govern and load share functions to perform over a full range of operating points and inhibit the
synchronizer from reaching full expected range. Conversely, setting the bias range too wide can make
a system unstable as it will be overly responsive to analog input noise or other minor deviations in the
bias signal, and analog closed loop gain would be unnecessarily high.
Failure of a module should be considered when setting bias signals. Generally the system should be
designed so that failure of a module should result in a bias signal of 0.
7.1.2.1 Configuration
Use InPower to configure the “Genset Application Type” as “Standalone” as shown in Figure 33.
FIGURE 34. CONFIGURING ANALOG INPUTS AS SPEED AND VOLTAGE BIAS INPUTS
Select “Speed Bias Input Scaling Table” as shown in Figure 35. The Voltage Bias Input Scaling Table
is also shown on Figure 35.
FIGURE 35. SELECTING SPEED AND VOLTAGE BIAS INPUT SCALING TABLES
The scaling table allows the operator to map the analog input voltage to a percent adjustment of the
frequency or voltage setting. Figure 36 is an example of a voltage bias scaling table.
For example, a speed bias of +5% would increase the frequency setpoint by 5% (to 63 Hz for a 60 Hz
system). The governor will increase fuel in an attempt to raise the frequency. When a generator is
paralleled to another source its ability to unilaterally change frequency is limited, so the increase in
fuel will result in an increase in power output. As the power output gets closer to the desired output
the bias signal should decrease, and eventually settle at 0. Similarly, a negative bias signal should be
used when it is desired to decrease the power output. A negative bias signal will lower the frequency
setpoint resulting in a decreased power output. As the power output gets closer to the desired level,
the bias signal should increase and eventually settle at 0.
FIGURE 38. GRAPHICAL REPRESENTATION OF SPEED BIAS SCALING TABLE WITH CLAMPS
The clamping function as illustrated in Figure 38 is one means to allow the bias signal to be very
responsive for smaller bias signals to but have a limited response to larger signals to prevent the
system from going unstable. The clamp can help dampen the response of the generator set if it is
receiving signals from an unstable paralleling module.
The voltage bias signals are inherently clamped at +10% and -40% even if the scaling table indicates
that they will go beyond those values.
When using the 3.3 control as a paralleling module to bias a 3rd party generator control there are four
additional settings that need to be set in InPower:
• ECM Can Enable must be set to “Disabled”
• AVR Enable must be set to “Disabled”
• Starter Owner must be set to “ECS”
• External Bias Commands Enable must be set to “Enabled”
When using bias I/O for paralleling validating system performance becomes more critical and more
complicated as there is an additional potential source of instability. Unlike load sharing and load
governing consisting of standard products, a system of dissimilar products has not been validated.
Some key steps must be considered:
1. If there is any instability set the bias range to the minimum possible value
2. Confirm that V/Hz characteristics for all generator sets are identical
3. Adjust governor and/or AVR gains as necessary for stability and performance. It is not unusual to
have to reduce gains to ensure stability
Refer to the commissioning and troubleshooting section (Appendix 1) for more information. A full
system validation is required in these systems.
PowerCommand voltage and speed bias outputs are disabled and isolated until the control is
operational, so that the system doesn’t drive the generator set to an unknown state at power up.
FIGURE 42. CONTROL INTERCONNECTIONS BETWEEN A 3.3 AND CAT EMCP 2.2.
The 3rd party generator control is still managing all of the generator protection functions and will shut
the generator set down under fault conditions. A common alarm or shutdown output from the 3rd party
generator control should be wired to a configurable customer input to the paralleling control so that the
paralleling control can raise the alarm to the building management system and be coordinated to
operate the breaker as appropriate. An operator will have to reset the fault at the 3rd party generator
control to re-start the generator set.
There should be one system fault reset from the user's perspective. The fault reset must work with
both control systems. To ensure proper "system" shutdowns, the generator set control needs to tell
the paralleling control to shut down. The paralleling control also needs to tell the generator set control
to shut down if the paralleling control detects a shutdown condition. When doing this, the generator
control needs to coordinate a shutdown command to the paralleling control and the paralleling control
needs to send a common shutdown to the generator control.
This adds complexity to performing a system reset. The system reset needs to tell the generator
control and the paralleling control to reset, and it needs to interrupt the "common shutdown"
communications between the two controls.
Table of Contents
Figure 43. Load Share kW and kVAR Balance .................................................................................... 65
A.0 Commissioning
Before attempting to close the paralleling breaker to another source:
1. Verify all parameter settings
• Confirm that reverse kW and reverse kVAR are within the capabilities of the
generator set. Reverse kW should be less than the Regenerative power rating
of the engine. Reverse kVAR should be within the limits shown on the
alternator capability curve. Consult the factory to obtain the alternator
capability curve.
• Confirm that V/Hz settings are identical for all paralleled generator sets.
• Consult the factory if it is intended to parallel generator sets with a continuous
leading power factor load or if paralleling with a 3rd party control system.
• Follow the paralleling setup screens in InPower.
2. Before attempting to parallel generator sets, it is important to confirm that the
generator sets individually run stable under load. All generator sets have to run at
the same frequency and voltage, since the load share control is "proportional" only
in Isochronous mode.
• To make all the generator sets have the same frequency and voltage, the
Genset Application Type must first be set to "Standalone" mode for each
generator set being paralleled.
3. While running on a dead bus (no other source is available and load is disconnected)
confirm that generator set circuit breaker operation and feedback works properly.
4. If the application involves utility paralleling confirm that the utility circuit breaker
feedback works properly. If the application involves operation of the utility breaker
(PTC operation) confirm utility circuit breaker operation with no utility source present
and no load connected. This involves making manual connections to the TB5
terminal on the PowerCommand circuit board to confirm that the breaker control is
wired correctly and the breaker functions. As an alternative, if the breaker can be
racked out into test mode the PowerCommand control can directly control the
breaker without opening or closing on to utility voltage.
5. For each generator set confirm load share line voltages are correct while running in
isolated bus mode but not paralleled to another genset. Load share line voltages
should measure ~0.25V at no load and ~2.5V at full load. Note that this assumes
that the MCM3320 Master Synchronizer is not driving the load share lines. Confirm
that polarity is correct on each generator set. Load share line polarity should match
for each generator set, and the MCM3320 (if applicable).
6. Calibrate PowerCommand AC measurements.
7. Confirm that the dead bus close arbitration does not have an open circuit. This can
be done with a resistance check across the entire circuit.
8. With the generator circuit breaker closed to a dead bus, confirm that generator set
voltage phase A matches bus voltage phase A and the phase difference between
generator set phase A and bus phase A is 0 degrees. Confirm for phase B and C as
well.
9. When one generator set is closed to the paralleling bus confirm that each of the
other generator sets sees that the paralleling bus is live and measures voltage
correctly.
10. Confirm that the each generator set has a different master first start backup time
programmed. One simple method for doing this is to assign each generator set a
number and set up a backup time in seconds equal to 10 times that number (ie
genset 1 would have a backup time of 10 seconds, genset 2 would have a backup
time of 20 seconds, etc.)
11. Note that it is important to verify that sensing leads are matched correctly to power
conductors before closing a paralleling breaker. During commissioning or if power
conductors or sensing leads are changed in any way It is always recommended to
verify the power leads are truly synchronized independent of the control sync check
function before closing a breaker. Prevent breaker closure by either running the
system in manual mode or by activating the breaker inhibit input when in auto mode.
12. Verify synchronizer performance as discussed in the synchronizer section of this
document.
13. All setup assumes no leading PF operation. If leading PF operation is needed,
consult the factory.
Table of Contents
Table of Contents
A final challenge was that the EMCP4.4 control was designed to control its own generator
set mounted paralleling breaker and we were not able to disable that feature. The 3.3
was designed to control a generator paralleling breaker in the switchgear. The typical
sequence of operations is that when the generator set senses the speed and voltage
have reached nominal levels it is ready to load and begins dead bus arbitration and
closes its paralleling breaker when it senses a dead paralleling bus, or starts
synchronizing if it senses that the bus is live. The difficulty here is that we had two
controls trying to control two breakers to execute the same function. The solution was to
move the 3.3 voltage sensing to the load side of the Cat generator set mounted
paralleling breaker and allow the Cat control to close its breaker as soon as it was ready
to load (assuming the bus is dead). The 3.3 control then will sense the generator set as
ready to load only after the generator set mounted breaker closed. At this point, the 3.3
controlled the paralleling breaker in the switchgear as it would in any typical paralleling
system.