Lawrence Berkeley National Laboratory

This report describes the tests conducted to determine the behavior of two Capstone 30 kW microturbines connected in parallel with some impedance between them. This test was meant to simulate the operation of two microturbines at nearby customer facilities. This arrangement also constitutes a simple microgrid. The goal of this test was to investigate if any voltage and power instabilities exist between the two microturbines. Two test sequences were conducted. The first test sequence operated two microturbine/load bank pairs using manual control of the microturbine and load bank setpoints. The second test sequence used the Capstone Load Following mode of operation to control generation levels of one of the microturbines. The two microturbine/load bank sets were connected together through a 300 foot long, four conductor, #12 cable so th at the impedance between them would cause up to a 5 percent voltage drop (depending on the load balance between the two sets). Data was collected from both Capstone microturbines, two power quality instruments and a power monitor. The tests showed there were no instabilities in the microturbines' voltage or power output as long as care was taken not to overload either unit. A voltage drop of almost 5 percent was observed between the two microturbines at the highest loadings. This "soft" connection between the two microturbines did not cause problems. Basic protective functions of the microturbines avoided unintentional islanded operation, but probably would not be sufficient or desirable for normal microgrid operations. Some simple automatic load sharing could be accomplished by using the Load Following mode of operation of the microturbine. This ability was demonstrated during the second set of tests. Load sharing works as long as all loads are kept within operating limits of the two microturbines. Use of the load following mode of the microturbine seemed to work fairly well. To improve the responsiveness of the load following, a faster pulse rate would need to be obtained from the kilowatt-hour meter for the expected loads. The faster pulse rate would allow a shorter averaging period in the microturbine which would make it respond more quickly. Revision of the microturbine operating software would be desirable so that both microturbines could share load and voltage regulation duties in the microgrid. Once these abilities were integrated into the microturbines, the functions would need to be verified in a set of lab tests and then checked in actual field operations. Additional protection functions would also need to be integrated into the microgrid so that a fault on the microgrid would not drop all load and generation.