Should a utility manage its system voltage if it will reduce its wholesale power bill? The Kootenai Electric Cooperative Inc. (KEC) management team asked this question after experiencing a significant increase in its wholesale power costs in 2010 as a result of rate changes. It found implementing a demand-response by voltage-reduction system (DR-by-VR) would reduce those costs. The system has saved KEC US$180,000 on its wholesale power bill in the first four months of operation. As a not-for-profit electric cooperative, these savings are passed on to customers in the form of lower electric rates.
Because of low wholesale demand charges the Pacific Northwest experienced prior to 2010, the implementation of demand-side management programs lagged significantly behind other regions in the U.S. After 2010 and the resulting demand charge increase, KEC experimented with water heater and home heating thermostat control, but found the systems were expensive to implement and did not have a high enough return on investment. The needs of participating customers also required a lot of staff time.
Energy conservation helps to reduce the utility’s overall demand, but it also reduces the utility’s energy sales and results in the need for rate increases to maintain operating margins. This results in a vicious cycle of eroding sales and increasing utility rates. For these reasons, KEC looked for other ways to help reduce its wholesale power bill without impacting its energy sales and considered voltage optimization.
Rates and Motivation
KEC serves 25,000 customers in the Coeur d’Alene, Idaho, U.S., area and has a peak annual load of 110 MW. The cooperative purchases most of its energy and all its capacity from the Bonneville Power Administration (BPA). Prior to 2010, KEC’s wholesale demand charges were below $3/kW because of surplus hydropower capacity at BPA. However, in 2010, things changed, and BPA no longer had a surplus of capacity in the system.
BPA changed its wholesale energy and capacity rate structures, implementing a tiered rate methodology for energy and capacity. Under the new tiered rate structure, each customer received a prorated share of the BPA billing. If the utility’s demand exceeded its allocation in a given month, it was assessed a significantly higher demand charge on the amount above its allocation.
Under this tiered rate structure, KEC is now subject to multiple types of demand charges. BPA’s PF-16 demand charge is directly related to the amount of the utility’s demand above its BPA monthly allocation. BPA’s NT-16 network transmission demand charge is directly related to the use of existing BPA transmission capacity during BPA’s transmission system peak load hour each month.
BPA’s PF-16 rate schedule monthly demand charge ranges from $6.20/kW to $11.47/kW, with late fall and winter months being the highest cost of the year. BPA’s network transmission demand charge is $2.04/kW. The result was a combined demand charge ranging monthly from a minimum of $8.24/kW to a maximum of $13.51/kW. BPA reviews the rates every two years and adjusts them accordingly, generally higher.
Voltage-based demand-response systems have been in use for years in the more urbanized areas of the Midwest and East Coast, primarily in population-dense areas with short distribution lines. The amount of load that can be shed varies by feeder and is related to the demographics of the load on each feeder, the length of the feeder and conductor type. Both Duke Energy and Alabama Power Co. implemented voltage-reduction control systems and found they could shave their peak demand by 1.6% to 2.7%, while staying within the service voltage range dictated by American National Standards Institute (ANSI) C84.1. According to available literature, the utilities could achieve between 1% and 5% demand reductions.
Electric cooperatives historically have avoided using voltage-control-based demand-response systems because of their long distribution feeders and the challenges they experience maintaining end-of-line voltage. It is not unusual for a rural power distribution feeder to have one to two voltage regulator banks downstream from the substation. Until recently, setting up a communications system to talk to each voltage regulation bank was expensive. However, continued improvements in cellular network coverage have dramatically reduced the cost of communicating with devices such as regulator banks located in remote areas.
Likewise, monitoring end-of-line voltage at the customer’s meter base to ensure service voltages stayed within the boundaries dictated by ANSI C84.1 was a costly challenge. Utility automatic meter reading systems have changed that situation, reducing costs in a similar manner to how cellular reduced downline regulator control costs.
Unsure of the financial implications of implementing a DR-by-VR system, KEC developed a financial model to analyze the impact of DR-by-VR on its wholesale power bill. The model enabled the cooperative to determine how much it would save on its wholesale power annually based on the percent demand reduction achieved. The effect of shaving the utility’s demand from 1% to 5% was simulated and the net financial impact calculated. All of BPA’s tiered demand rate structures were modeled to capture the total net effect of shaving KEC’s peak demand.
For the purposes of the analysis, it was assumed DR-by-VR used both peak shaving and valley filling. The system would lower the distribution system voltage only during the system peak and raise system voltage during the lowest demand part of the day, offsetting the lost kilowatt-hour sales caused by the peak shaving efforts. The model also assumed KEC maintained its service voltages within regulatory limits.
The annual demand charge savings KEC could obtain at different demand-reduction percentages meant even a modest 2% demand reduction could save the utility a significant amount on its wholesale power bill. Based on the financial model, the range of savings KEC could expect to see using voltage optimization (peak shaving and valley filling by voltage control) ranged from $104,963 with a 1% demand reduction to $475,790 with a 5% demand reduction.
To see what actual peak demand reduction KEC could expect to obtain, the utility conducted a field test at its Julia Street substation, which has short urban feeders. By dropping the voltage down to 116 V on a 120-V base, the demand dropped 5.2%. The result was enough to compel KEC’s management to implement the DR-by-VR system.
KEC already had a supervisory control and data acquisition (SCADA) system installed at each of its 13 substations, which it leveraged for its voltage control system. The cooperative also leveraged its Aclara TWACS power-line-carrier-based advanced metering infrastructure (AMI) system to obtain end-of-line voltage readings on each feeder at its customer meter bases. While many utilities only implement a voltage control system at the substation bus level, to gain the most benefit from its system, KEC felt it needed to control the midline voltage regulators, as well.
In order to control the midline regulators, the cooperative decided on an architecture using cell modems and a microprocessor-based controller installed at each midline voltage regulator. A feedback loop was created by using voltage readings from the end-of-line meter locations back to the DR-by-VR system software. The DR-by-VR control software then would change the voltage regulator set points to accomplish the necessary voltage reductions along each feeder line section while keeping the lowest customer’s voltage level within the allowable limits.
KEC worked with SCADA Nexus, its SCADA integrator, to provide the DR-by-VR data acquisition and control system for the substation and midline voltage regulators. SCADA Nexus also developed a feedback control software add-on for the system. It developed the software interface to the Aclara TWACS AMI system that initiates on-demand voltage reads from specific utility meters at customers’ homes located at the end of each voltage regulator control zone. KEC has established more than 100 control zones, with an average of three meters surveyed in each zone per phase.
The system enables KEC to track the amount of energy sales lost during voltage-reduction events and recapture only the lost energy sales by implementing valley filling. This is accomplished by raising up the voltage near the upper range of ANSI C84.1 Range A during a period of off-peak system operation.
As KEC is a not-for-profit organization, it was important to ensure valley filling only recaptured the kilowatt-hour sales lost during the voltage-reduction events. The software developed institutes a net-zero accounting for the valley-filling automatic operation period accomplishing this goal. KEC wanted to ensure the system’s operation did not directly affect customers’ monthly bills. Rather, customers would benefit, because the savings KEC obtained on its wholesale power bill would be passed on to them in the form of lower rates.
The DR-by-VR system was made operational, and the utility ran its first systemwide test on Sept. 15, 2016. During the test, the system was used both to reduce the system peak demand and to flatten the utility’s load profile during the lightly loaded time of the day. In demand-reduction mode, the system voltage was lowered to 114 V on a 120-V base. During valley-filling mode, the system voltage was raised up to 126 V to recapture lost energy sales.
To date, KEC has received only three voltage complaints out of 25,000 metered customers. After investigation, two of the complaints were associated with owners of motor drives who had their trip-out set points set above the lowest service voltage range values allowed by regulations. The third complaint was from an AM radio talk show host who had a broadcast studio set up in his home. All three complaints were resolved satisfactorily by educating the customers on the voltage service ranges dictated by ANSI C84.1. The owners of the motor drives changed the trip-out settings on their motor controls, and the radio talk show announcer was happy his electric cooperative was working to control costs.
During the three-month test period, KEC discovered some tuning of the system was required to stay within ANSI C84.1 lower limits, as the system tended to overshoot the lower limit a little before correcting and bringing the system voltage back into the allowable range. To compensate for this, the initial lower set point was raised up a couple volts to 116 V, instead of going directly down to 114 V during the initial control adjustment.
Now on Auto
After three months of manual testing, the DR-by-VR system was switched to auto and now actively manages KEC’s peak demand. During the first four months of operation (from December 2016 through March 2017), the DR-by-VR voltage optimization system saved KEC $180,000 on its wholesale power bill demand charges. At this rate, the system will pay for itself in less than two years. The savings realized by the voltage-based demand-response system exceeded KEC management’s expectations and will benefit customers by helping the cooperative to keep its rates low. ♦
Shawn Dolan is manager of engineering for Kootenai Electric Cooperative, located in Hayden, Idaho. He holds BSEE and MSEE degrees from Montana State University in Bozeman, Montana, and is a registered professional engineer in the states of Oregon, Washington, Idaho and Montana. He has 28 years of experience in the electrical utility industry, including 18 years with KEC.
Scott Hamilton is founder and president of Resource Associates International Inc., based in Spokane, Washington. He has a BSEE degree from the University of Idaho and is a registered professional engineer in the state of Washington. After graduating, he joined Washington Water & Power Co. (now Avista Corp.) in Spokane as a staff engineer, where he worked in various design and mid-level management positions. While at Washington Water & Power, he designed SCADA system upgrades, new substation systems and control systems for power plants.