KCP&L reaps bottom-line benefits from voltage control, which is now a normal operating practice.
In early 2006, KCP&L executives looked for ways to reduce system peak demand. The distribution automation engineering group proposed a voltage reduction option called dynamic voltage control (DVC). After a successful proof-of-concept pilot in 2006, DVC rapidly gained momentum. The original three-year plan was compressed to two years.
KCP&L had nearly 1,000 automated two-way communicating capacitors installed in the Kansas City metropolitan area prior to the launch of DVC. Load in the area was 90% of the total peak, while the geographic area was less than 20% of the total service area.
KCP&L's metropolitan circuits are relatively short. Voltage regulation is predominately accomplished with three-phase substation load-tap changers (LTCs) and controls working with automated distribution capacitors that switch to support system voltage and voltage-ampere-reactive needs. Very few distribution voltage line regulators are installed on metropolitan circuits. However, metropolitan circuits were ideal candidates for lowering demand with a voltage reduction scheme. The existing base of automated capacitors became an economic enabler for the DVC project. All that was needed to implement the project was the incremental costs to automate the LTCs through the energy management system (EMS).
KCP&L had numerous older substation LTC controls in its substations in 2006. Many of the existing LTC controls were difficult to calibrate and maintain. The project enabled the upgrade of these older controls to intelligent electronic devices (IEDs) that could use distributed network protocol (DNP).
Automation engineers collaborated with substation engineers, relay engineers and real-time system engineers to optimize the best blend of technology and economics. They decided to integrate the substation LTC IED into the remote terminal unit (RTU). For KCP&L, one LTC control typically regulates substation switchgear that serves up to four distribution circuits. All of this was integrated into the EMS.
The project was designed to address several objectives:
Improve substation and distribution voltage regulation
Enable additional system monitoring
Project Pilot Phase
Lower system peak demand
Improve management and implementation LTC settings
Implement remote LTC control
Provide a platform to better support future substation automation.
The LTC streams DNP messages to the RTU. Engineers encouraged robust use of DNP points. Substation engineers took the design a step further by building various RTU algorithms to enhance performance. Select employees could change LTC control setpoints and input specified voltage reduction values.
Users initiate voltage reduction at each of 203 buses by sending a command to activate the reduced voltage settings for each substation bus. There are three standardized voltage reduction values for every bus, each having target values for demand reduction. Standard voltage reduction percentages are 1.5%, 2.5% and 3.5%. However, if reduction is too aggressive for a given bus, the target percentage may be customized for that bus.
In 2006, DVC was installed on two substation buses, serving eight circuits, to prove the concept worked and to perform statistical testing. Engineers studied the relationship between the percent of megawatt reduction for each unit of voltage reduction implemented. Interaction between changes in voltage versus the effect on megawatt load was studied at both the circuit and bus level. Revenue meters were installed to monitor the main breaker load of the distribution switchgear and one feeder breaker load. Data was captured and retrieved through a substation data recorder every five seconds to enable historical trending. The data polled at each circuit and bus were megawatt, megavolt-ampere-reactive, amps and volts on each phase, power factor, ambient temperature, and LTC controls and inputs.
Studies showed megawatt reduction per unit of voltage reduction by implementing and releasing voltage reduction control for a given circuit on a given day. KCP&L also reviewed industry studies as benchmarks. In addition, engineers analyzed the composition of residential versus commercial load for each test circuit. Results varied and were unique for each circuit based on customer loads. However, KCP&L determined system aggregate results of 0.92% megawatt reduction for each 1.0% voltage reduction during summer peak loading.
Early Practical Problems
The aggressive deployment rate plus the new technology placed the project completion date at risk. A KCP&L project manager coordinated regular meetings to keep resources informed of critical paths to accomplish the project in four phases:
Phase I - 60 buses in service by end of 2006
Phase II - 100 total buses in service by summer of 2007
Phase III - 140 total buses in service by end of 2007
Practical Implementation by Dispatchers
Phase IV - 203 total buses in service by end of 2008.
The equipment was ordered and staged to facilitate this schedule.
EMS engineers and technicians were responsible for building related operator screens. EMS screen developers interviewed engineers and dispatchers to determine user-friendly EMS screens. They also worked long hours with field crews to check each DNP point at each IED and RTU.
The DVC screen shows both analog input values and pending analog output values for the setting changes desired:
Resistance (used to raise and lower the bus voltage as load changes)
Voltage band center
Enjoying Real Benefits
DVC voltage reduction percentage setpoints (three for each LTC).
Prior engineering practices for substation LTC controls called for a standard voltage center point with a bandwidth of ±1.5 V. However, when these settings were applied, customer low-voltage complaints increased. Additional investigation showed the LTC center band was typically set in the field at 1.0 V higher than the standard by substation technicians based on operating experience. After taking this practical input into account, a new standard for voltage center band was established. With DVC, engineers could monitor and change LTC setpoints from their secure EMS desktop application, eliminating truck rolls and customer complaints.
Typically, LTC controls were an earlier vintage design that required mechanical adjustments to the control. Such controls were difficult to calibrate for accurate settings. With the new IEDs, settings are established, displayed and maintained through the EMS. Rogue field changes are automatically rejected by the system.
However, engineers noticed remote settings through the EMS did not always stay at the set value. Substation engineers worked diligently to discover the root cause. KCP&L was in the process of moving from an older EMS that used legacy CDC type II protocol to a new EMS platform that uses DNP. The older EMS master would poll the tap-changer controller IED with DNP, but the substation RTU converted it to a CDC type II message, linking the EMS to the IED. CDC protocol did not have secure analog output features (such as “select before operate”) typically used with modern controls.
CDC had a limitation of only 16 analog output points per RTU. Therefore, the single RTU cabinet was configured with as many virtual RTUs as needed on separate RTU ports to achieve the required number of analog outputs. The substation RTU analog leased line was shared by all the RTUs at the substation to the EMS.
Sharing multiple RTUs on two-wire phone circuits caused the RTU to erroneously interpret what it received as a valid analog output command, changing the setpoints. Custom RTU programming provided a safeguard only allowing analog setpoint changes through a virtual remote/local switch. When the switch was set to remote through EMS command to the RTU, a special timer was programmed to limit setpoint changes within a 10-minute window. This prevented inadvertently leaving the station in the remote position. A digital alarm also monitored whether setpoints from the EMS were different values than in the IED control. Quick resolution of this technical issue was key to completing the project on schedule.
KCP&L's DVC implementation is designed for use during summer system peak hours. When ambient temperatures reach mid-90°F (32°C), all automated capacitors are designed to switch on to support peak loads. This establishes a steady condition between distribution capacitors and LTCs. Dispatchers turn on voltage reduction based on a specified target group. Typically, KCP&L activates the 2.5% voltage reduction for DVC.
Once DVC is implemented, dispatch supervisors use tools to monitor voltage at all automated capacitors. Each capacitor has a voltage sensor installed on the A phase to correspond to voltage monitoring at the substation LTC control that also is monitoring and controlling this same phase.
Dispatchers receive and monitor alarms when voltage drops below set values. They also track circuits that are temporarily reconfigured to ensure no voltage violations exist at capacitors on these circuits. If voltage at a capacitor bank falls below the threshold, dispatchers have the option of issuing commands to change the bus setting to the 1.5% voltage reduction target or releasing voltage reduction completely for that substation bus.
Each year, planning engineers review LTC settings for the metropolitan system. In addition, DVC enables KCP&L to respond to customer voltage complaints caused by temporary system reconfiguration.
KCP&L dispatchers and linemen expected a step function of new voltage complaints when DVC was active. Thankfully, this has not been the case. Should a problem arise, it is typically due to some other installation or operational problem. For example, one customer complained of low voltage during peak loading conditions when the air conditioner was running. A field check showed the customer was served from an overhead distribution transformer with long secondary mains and also a long service drop. After the transformer's tap position was reset, voltage improved and there were no problems for this customer, even during DVC events.
In another instance, a substation bus automatically transferred to a tie position caused by substation transformer failure. The LTC then raised the voltage to an unacceptably high value. Dispatchers immediately used the remote LTC feature to lower the voltage to an acceptable level. Thereafter, planning engineers provided and used DVC to promptly implement a new temporary LTC setting by remote control to restore proper automatic voltage regulation at the substation.
DVC played a strategic role in demand reduction in the summer of 2011 for unexpected reasons. Electricity in the Midwest was at a premium because of power shortages in Texas and other states. The 2011 summer was much hotter than normal, with many days over 100°F (38°C). KCP&L used DVC successfully, regularly and economically to lower system peak.
The KCP&L DVC design is innovative, practical and economical. There have been fewer distribution voltage complaints since the DVC implementation, mostly due to the fact the substation LTC controls are now upgraded and can be managed and controlled remotely. KCP&L installed the DVC system for an average cost of US$8,500 per circuit.
Dispatchers now can use remote control to immediately switch an LTC to address voltages due to emergency system reconfiguration. Settings are now effectively managed and monitored in addition to being more precise. Operations and maintenance costs have been reduced. Periodic field visits to check settings and calibrate older electromechanical LTCs have been eliminated by the use of electronic LTC controls with the EMS interface. In addition, KCP&L has upgraded various RTUs to enable future substation automation applications.
This DVC system has been successful and sustainable because it works automatically with little database or system maintenance. The use of DVC in KCP&L's dispatch center is now standard procedure and operates seamlessly with normal operating practices. Additional studies show DVC has provided demand reduction near 50 MW, which is very close to the anticipated demand reduction. Furthermore, similar engineering studies showed a release of at least 30 MVAR during DVC implementation that provides economic alternatives for voltage-ampere-reactive support versus incremental installation of distribution line capacitors.
KCP&L has realized all its goals for the DVC program and is now enjoying real benefits of this program.
Carl R. Goeckeler (email@example.com) is the lead distribution automation engineer at KCP&L. He received his BSEE degree in power systems from the Missouri University of Science and Technology in 1975. He has spent 36 years at KCP&L in design and operations in transmission, substations and distribution. He initiated KCP&L's power-quality program and is the author of a syndicated power quality training publication used by more than 20 utilities. Under Goeckeler's direction, KCP&L has won five national awards for its distribution automation projects. He is a registered professional engineer.
Beckwith Electric Co. www.beckwithelectric.com
Kansas City Power & Light www.kcpl.com