SatCon Technology Corporation (United States)

Initiated in 2008, the Solar Energy Grid Integration Systems (SEGIS) program is a partnership involving the U.S. DOE, Sandia National Laboratories, private sector companies, electric utilities, and universities. Projects supported under the program have focused on the complete-system development of solar technologies, with the dual goal of expanding utility-scale penetration and addressing new challenges of connecting large-scale solar installations in higher penetrations to the electric grid. The Florida Solar Energy Center (FSEC), its partners, and Sandia National Laboratories have successfully collaborated to complete the work under the third and final stage of the SEGIS initiative. The SEGIS program was a three-year, three-stage project that include conceptual design and market analysis in Stage 1, prototype development and testing in Stage 2, and moving toward commercialization in Stage 3. Under this program, the FSEC SEGIS team developed a comprehensive vision that has guided technology development that sets one methodology for merging photovoltaic (PV) and smart-grid technologies. The FSEC team's objective in the SEGIS project is to remove barriers to large-scale general integration of PV and to enhance the value proposition of photovoltaic energy by enabling PV to act as much as possible as if it were at the very least equivalent to a conventional utility power plant. It was immediately apparent that the advanced power electronics of these advanced inverters will go far beyond conventional power plants, making high penetrations of PV not just acceptable, but desirable. This report summarizes a three-year effort to develop, validate and commercialize Grid-Smart Inverters for wider photovoltaic utilization, particularly in the utility sector.

Dramatic improvements in Grid power quality require fundamental changes to the physical layer of the Grid, where our electric power is generated, transported and utilized. The addition of “Smarts” to the Grid is essential; however, while it is a necessary condition, it is insufficient to provide the digital quality power that our loads increasingly demand. The fundamental problem of the Grid is not lack of intelligence, but lack of speed. Bulk power sources are remote from loads, and existing controls are applied to clumsy actuators and switchgear. The Future Grid, described here, requires a minimum penetration of generation coupled through electronics. This can easily be achieved through Grid-connected renewables, ideally highly distributed, with autonomous controls for transient events and centralized regional controls for slower time scales. Partial penetration of fast response generation enables performance which can achieve substantially higher quality of power. An additional requirement for a truly robust Grid is either a very high level of high bandwidth control of sources and loads or high-speed fault-limiting and high-speed valves in the Transmission and Distribution (T&D) system, to contain damage and prevent cascading blackouts.

When designing and building power systems that contain power electronic switching sources and loads, system integrators must consider the frequency-dependent impedance characteristics at an interface to ensure system stability. Stability criteria have been developed in terms of source and load impedance, and it is often necessary to measure system impedance through experiments. Traditional injection-based impedance measurement techniques require multiple online testing that lead to many disadvantages, including prolonged test time, operating point variations, and impedance values at limited frequency points. The impedance identification method proposed in this paper greatly reduces online testing time by modeling the system with recurrent neural networks with adequate accuracy. The recurrent networks are trained with measured signals from the system with only one stimulus injection per frequency decade. The measurement and identification processes are developed, and the effectiveness of this new technique is demonstrated by simulation and laboratory tests.

This paper presents a predictive based current limiting controller for a boost converter that was developed to charge an electric vehicle battery from the vehicle propulsion bus in an automotive bi-directional conversion system. To meet aggressive power density requirements, it was necessary to operate the three-stage interleaved converter magnetics near the edge of saturation of the boost inductors. Under large load transients the converter control needed to operate the converter as a current source to prevent hard saturation of the magnetics. The predictive limiter was used to achieve maximum slew rate while limiting the current. The implementation used background model calculation and lookup tables to minimize real-time computation and associated delay. The paper presents details of the current limiter design for a boost converter; its implementation into a DSP based controller and experimental results. Further enhancements and other applications are also discussed.

This paper presents the design and experimental verification of a 6 kW interior permanent magnet (IPM) synchronous machine intended for an automotive direct-drive starter/alternator application. The machine was designed using a saturable lumped-parameter magnetic circuit model in combination with a Monte Carlo optimization process that minimized the machine-plus-converter cost. An experimental IPM machine has been constructed based on the resulting design specifications. Laboratory tests have confirmed the accuracy of the analytical models for predicting the q-axis inductance L/sub q/ (including saturation effects) and the torque production characteristics, but discrepancies between the predicted and measured d-axis inductance L/sub d/ were revealed. The impact of these differences on machine performance is discussed, as well as potential adjustments in the IPM analytical model to improve the performance of future machines.

Growing automotive auxiliary electric power requirements have resulted in considerable interest in an integrated starter/alternator (ISA). Initial efforts in this area have been focussed on a direct-drive ISA configuration which offers an appealing degree of mechanical simplification. However, an offset-coupled ISA configuration similar to that used by the conventional alternator has potential for providing a better system solution in some applications, including lower total system cost. This paper first describes the design, construction and dynamometer testing of a concept demonstrator of an offset-coupled ISA using an interior permanent magnet (IPM) synchronous machine. Next, a nonlinear lumped-circuit analysis tool is used to investigate optimized offset-coupled IPM ISA drive designs. Finally, comparisons are presented between an existing direct-drive IPM ISA design and the optimized offset-coupled designs.

This paper discusses mechanical design that are particular to conventionally (i.e., radially) laminated rotors of interior permanent-magnet synchronous machines. Focus is placed on applications where the radial forces due to high-speed operation are the major mechanically limiting design factor. Proper design of the lamination bridges, or ribs, at the rotor outer diameter is explained in terms of the both material considerations and electromagnetic performance impact. The tradeoff of complexity versus performance associated with using strengthening ribs in the magnet cavities is discussed. The sensitivity of the mechanical design limitations to the rotor-shaft mounting mechanism is also highlighted. These effects are then analyzed using finite element analysis for a 150 Nm/6 kW integrated starter/alternator designed for operation up to 6000 RPM with an annular rotor to accommodate a torque converter or clutch assembly. This example demonstrates that it is possible to significantly improve the rotor's structural integrity using the techniques described in this paper with only a very modest impact on the projected machine drive cost.

The authors designed and implemented a magnetically levitated stage with large planar motion capability. This planar magnetic levitator employs four novel permanent-magnet linear motors. Each motor generates vertical force for suspension against gravity, as well as a horizontal force for drive. These linear levitation motors can be used as building blocks in the general class of multi-degree-of-freedom motion stages. In this paper, they discuss the electromechanical modeling and real-time vector control of such a permanent-magnet levitator. They describe the dynamics in a dq frame introduced to decouple the forces acting on the magnetically levitated moving part, namely, the platen. A transformation similar to the Blondel-Park transformation is derived for commutation of the stator phase currents. They provide test results on step responses of the magnetically levitated stage. They show 5 nm RMS positioning noise in x and y, which demonstrates the applicability of such stages in the next-generation photolithography in semiconductor manufacturing.

SatCon Technology Corporation has demonstrated excellent vibration reduction performance using active control on the JPL Stirling-cycle cryocooler testbed. This paper addresses the use of classical narrow-band feedback control to meet the cryocooler vibration specifications using one cryocooler in a self-cancellation configuration. Similar vibration reduction performance was obtained using a cryocooler back-to-back configuration by actively controlling a reaction mass actuator that was used to mimic the second cooler.

Long-term insulin secretion was investigated in seven biohybrid capillary devices seeded with canine islets. Approximately 50,000 islets could be isolated from a single canine pancreas using collagenase digestion in conjunction with the recently described Velcro technique. Devices were perfused with tissue culture medium 199 containing 300 mg/dl glucose. Insulin secretion fell during the first 1-2 days of culture from approximately 18 to 6 U/day. After 3-4 wk of perfusion, however, there was a gradual rise in insulin output that reached greater than 15 U/day after 7 wk. Insulin output eventually stabilized at 18-20 U/day. Two of these devices were studied for 80 and 86 days, respectively, and continued to secrete these same amounts of insulin. Glucose-induced insulin release was studied after 1 and 4 wk of culture and was well preserved.