Austin Energy, the nation’s third largest municipality, aims for 65 percent renewable generation by 2027. This presents both operational and economic puzzles. One of the biggest is how to measure—and optimize—the total value of a system with such a diverse mix of distributed energy resources (DER). It’s one thing to calculate the levelized cost of energy (LCOE) for a single asset or system of assets like storage or solar. It’s another thing completely to do so for an entire utility enterprise with a portfolio of DER owned or operated by a variety of stakeholders in complex, interconnected, and expanding ways.
That’s exactly what the Austin SHINES project set out to do, receiving a $4.3 million grant in 2016 from the U.S. Department of Energy Solar Energy Technologies Office (SETO) to establish a working business model for DER optimization for grid, commercial, and residential applications. The project would build repeatable methodologies for designing and operating energy storage and solar photovoltaics (PV) on a grid. It would also define a “system LCOE to serve load” metric that encompasses the holistic costs and benefits of all combined resources.
Proving Ground with a Diverse Set of Assets
Austin SHINES provides a microcosm of that future grid. Here, on a smaller scale, the utility can experiment and explore how best to maximize its use of renewables to meet its future goals. These include attaining:
- 10 MW battery storage by 2025 and 30 MW thermal energy storage by 2027
- 750 MW utility-scale solar and 200 MW of local solar online, including 100 MW customer-sited PV by 2025
- 65 percent of customer load offset with renewable resources by 2027
- 1,000 MW of savings from energy efficiency and demand response by 2027
- Net-zero community-wide greenhouse gases by 2050
All these goals would be subject to meeting affordability goals as well, keeping rate increases to less than 2 percent a year and keeping utility rates in the lower 50th percentile in the state.
Over the last two years, Austin Energy assembled DER with a diverse team of vendor and community partners. Resources include two utility-scale energy storage systems (ESS), several customer-sited ESS at residential and commercial properties, smart inverters, real-time data feeds, a distributed energy resource optimizer, and a vehicle-to-grid (electric vehicle) component. Altogether, the project came to include more than 5 MW total—with resources on both sides of the meter.
Austin SHINES Resources
The size and variety of its resources allow the utility to explore, test, and evaluate what works best. Its resources include:
- Utility scale energy storage + PV
- 2.5 MW at La Loma Community Solar Farm
- 1.5 MW / 3 MWh Li-Ion Battery Storage at Kingsbery Substation
- 1.5 MW / 2.5 MWh Li-Ion Battery Storage at Mueller Substation
- Commercial energy storage + PV
- Aggregated storage installations at three sites
- One 18 kW / 36 kWh Li-Ion Battery Storage installation
- Two 72 kW / 144 kWh Li-Ion Battery Storage installations
- All sites have existing solar (300+ kW)
- Residential energy storage + PV
- Aggregated storage installations at six homes (10 kWh each)
- Each with existing rooftop solar
- Utility-Controlled Solar via Smart Inverters at twelve homes
- Autonomously Controlled Smart Inverters at six homes
Data is emerging from the project, now in full deployment, to inform the utility’s overall grid planning to reach its aggressive renewables goals.
Making the Operations-Economics Connection
One of the key steps to gaining a unified system LCOE, naturally enough, is unifying system controls across a utility enterprise. Without the ability to bring data together to inform comprehensive analysis of all available assets, it is not possible to consider all resources and find the most cost-effective overall solution. Second, and just as important, is equipping multiple systems to act intelligently to dispatch those assets that best contribute to the total system LCOE. Without this ability, a utility might know which resources to deploy at precisely which times and places but would not be able to do it.
To achieve these holistic capabilities, Austin SHINES designed layers (or levels) of intelligent controls, both distributed and centralized. The utility worked together with Doosan GridTech to design the solution around two key software systems. Doosan GridTech’s control software, Distributed Energy Resource Optimizer™ (DERO™) works at the centralized level, while Doosan GridTech Intelligent Controller™ (DG-IC™) works on a distributed level—along with additional behind-the-meter (BTM) controllers that bring in customer data. Together they provide a distributed energy resources management system (DERMS) that leaves no asset out, including those on the customer side of the meter (see Fig. 1).
Holistic results begin with the distributed controllers. They are, in a sense, the foot soldiers of the deployment—in the field, working with both a level of autonomy and in constant communication with central command. From behind the meter, ConnectDER™ units provide real-time data and control of some residential solar resources, while Pecan Street manages residential ESS and Stem manages commercial ESS.
DG-IC units manage other resources, including utility-scale resources at the Kingsbery and Mueller substations and several commercially sited BTM resources. Each DG-IC can provide real-time monitoring and control of 50 or more PCS-battery banks, solar PV and/or auxiliary devices within a customizable control system. All provide data back to inform central control intelligence through open standards that allow the utility to incorporate control units from different vendors and still feed up into the enterprise system. The approach overcomes the silos that often remain from legacy systems and equip the utility to scale up to meet future needs, potentially adding multiple fleets of aggregated resources.
More connections can mean more points of vulnerability, so Austin SHINES also includes layers of security protocols including local, remote, and automatic control functions equipping the utility to control and customize access. The design provides a three-tier alarm system to ensure safe operation. In a regional outage, the local controls could potentially also “island” grid sections to maintain local service.
In practice, all information from distributed intelligence is feeding continuously to DERO in the T&D control center via the open standards (see Fig. 2). These allow all resources, regardless of type, age, owner, or side of market, to contribute inputs to DERO where the holistic analysis takes place, unlocking new levels of actionable intelligence that deliver even deeper value.
In the Control Center
In addition to the grid data coming from the distributed intelligent control system, DERO also onboards data from five major sources within the control center. Market price data from the Electric Reliability Council of Texas (ERCOT) comes in through the utility’s energy desk (allowing DERO and its operators to learn from historic market pricing and to anticipate price movements through forecasts). Forecasts also come from clean power research, which projects the performance of various resources based on data compiled on typical outputs under different conditions. An operations historian (OSIsoft PI) feeds DERO data on past performance of its grid assets. It also accesses data from the utility’s advanced distribution management system (ADMS) and its SCADA.
Extracting Value from the Data—and Practice
With the multiple streams of information it receives, DERO runs analyses at the utility’s direction to support the five top use cases Austin SHINES has identified as most important to operational and economic value. These include energy arbitrage, peak load reduction, real-time price dispatch, congestion management and voltage management. Each of Austin SHINES’ diverse set of DER can help yield different results. And each use case is yielding distinct lessons on how best to extract that value.
Energy storage at commercial customers and at residential customers can reduce peak load. Figure 3 shows how intelligent controls work. Here, a commercial customer within the project applied insights from the data to stagger its use of energy-hungry HVAC systems and additionally relied on discharge of batteries during local peak hours, while coordinating with DERO for use of the batteries to reduce the utility peak. The tactic lowered the need for energy, reducing demand charges and transmission charges, benefiting both the customer and utility.
With the diverse sets of data coming into DERO, Austin SHINES can direct both utility resources and BTM customer resources to participate in day-ahead energy arbitrage and real-time price dispatch. There is no one-size-fits-all approach to this. DERO’s ongoing analysis provides active, tactical direction on the best ways to realize economic value from price variations in each situation. Currently, the storage deployments at Kingsbery and Mueller substations, as well
as the aggregated commercial ESS and residential ESS, all participate in energy arbitrage.
Power Quality Improvement
Solar PV sites are the only resources not currently participating in energy arbitrage. But they alone are engaged in voltage support, which aims to enhance value by reducing losses and helping to increase solar penetration.
DG-IC can address other power quality issues, too. With prioritizable operating modes for real and reactive power, it provides power smoothing and frequency response that increase the impact of energy storage systems. All of this can help make the integration of renewables cost-effective while upholding power quality and reliability.
Most of the DER are also part of congestion management efforts as well, working to enhance value by increasing local grid reliability. DERO can track congestion events, such as when solar peak production is bringing large amounts of reverse power flow onto the grid. Through daily use and practice the utility is extracting lessons and insights on the best ways to use ESS to reduce congestion.
The DERO screenshot from events on January 4 and 5, 2019, shows operators exploring changes in current limits to allow congestion relief (see Fig. 4). The first arrow shows preset limits that, when crossed, result in full action from the fleet. The second arrow shows how by raising limits the next day, operators could allow DERO to co-optimize other value opportunities with congestion events.
Congestion Management in Use
Finding the best tactics and techniques for different scenarios using DERO is a core aim of the Austin SHINES project. Here, the activities revealed the best tactics for maximizing the value in relieving congestion.
Putting It All Together
Each of the different use cases that Austin SHINES is exploring relies on DERO’s ability to analyze multiple types of information continuously and automatically, as well as to provide analysis on demand. The speed and power of the processing yield actionable insights that actively guide daily decisions. Taken as a whole, these actions contribute to a measurably optimized grid with a total system-wide levelized cost of energy.
How can a utility calculate this value? Austin SHINES has defined a formula:
The calculations are yielding a new way of understanding the combined value of holistically managed DER.
Performing analysis on the accumulated data from ongoing activity is confirming a sense that the cost per kWh is highest for all assets with no controls, lower when all assets gain autonomous controls, and lowest of all when the utility applies controls holistically. Early results on a single feeder at Austin SHINES show a 22 percent lower systemwide levelized cost of energy when the utility applied intelligent controls holistically (see Fig. 5).
The utility will continue monitoring, reviewing, and learning through fall 2019 when Austin SHINES expects to deliver its findings as part of the SETO Austin SHINES grant requirements. The insights gained on how best to holistically control and optimize value of complex systems, from both sides of the meter, should prove useful to an industry still learning as the penetration of DER continues to grow.