Navid Zargari (left) points out features of the PowerFlex 7000 pulse-width modulation AC drive. Source: Rockwell Automation.
Five years ago, electricians would get nervous if someone suggested eliminating the isolation transformer from a 1,000-hp, 2,300-V AC motor drive, particularly if the motor’s variable-speed drive draws a non-linear load. After all, transformers mitigate common mode voltage (CMV) stress that otherwise would lead to premature motor failure. But isolation transformers are space hogs and are expensive to install and maintain, and they reduce overall efficiency. If the harmonic needs of the power supply and the variable draw of the motor could be resolved by using harmonic-elimination techniques through switching-power semiconductors, reasoned electrical engineers at Rockwell Automation’s medium-voltage drives center in Cambridge, Canada, current and voltage waveforms would be near-sinusoidal, resolving CMV stress and eliminating the need for a transformer. Their work led to Direct-to-Drive technology and collaboration with Toronto’s Ryerson University to couple their drive with an integrated DC choke. The result is a significant reduction in the size and weight of the drives and a boost in Rockwell’s position in the market.
Medium-voltage motor AC drives typically range from 200-5,500 hp (150-4100 kW) if air cooled. One of the earliest commercial applications of the Rockwell Automation transformerless drive was in the food industry. La Union sugar mill in Guatemala installed a 1,000-hp, 1,200-rpm Reliance motor with an Allen-Bradley PowerFlex 7000 pulse-width modulation AC drive in November 2002. It replaced steam turbines driving mechanical equipment in a cogen system fueled by bagasse. The company realized a six-month payback.
In recognition of the research and development underlying the technology, the Ontario Ministry of Research and Innovation recently named Navid Zargari its Innovator of the Year for 2009. The award, which salutes breakthrough technology that results in commercially successful products, also salutes his collaborators: Steve Rizzo and Yuan Xiao of Rockwell and Bin Wu of Ryerson University. An electrical engineer, Zargari earned a bachelor’s degree in engineering at Tehran University, then studied at Montreal’s Concordia University, where he specialized in power electronics enroute to a PhD. He joined Rockwell Automation’s medium-voltage R&D center as a senior designer in 1994.
FE: What was the impetus for your work?
Zargari: If you look at a medium-voltage AC drive, the need is obvious to reduce the number of components and the power losses that occur in the drive system. Wouldn’t it be nice, we thought, to eliminate the transformer altogether? Before you can do that, it’s necessary to resolve variable power demand, harmonic oscillation and the negative impact of common-mode voltage on cable and motor insulation.
Our solution uses an active front-end rectifier technology with a common-mode DC choke. With conventional drives, the number of components adds a lot of engineering cost. In our design, harmonic mitigation is achieved through the active front end that uses the low-loss power semiconductor integrated to its gate driver. We call this symmetrical gate commutated thyristor technology. In a 2,400-V drive system, only 12 power semiconductors are needed.
FE: What other benefits does the technology deliver?
Zargari: There is a 0.5-1% loss in the transfer of power through a transformer. That is significant over time. By blocking common-mode voltage and mitigating motor neutral-to-ground voltage, there’s no need for an isolation transformer.
A transformer adds significantly to the footprint and weight of a drive system. A 1,250-hp/950-kW drive system with a transformer weighs 9,259 lbs. and fills 459 cubic feet. With Direct-to-Drive, the system weighs 2,976 lbs. and occupies 190 cubic feet.
FE: You were issued a patent for the multi-motor current source inverter drive in 2000. What was the next development step?
Zargari: We had the active front-end designed, but it took another 15 months to come up with the integrated choke. We started building samples in 2003. It took four attempts to get the design right. You have to do enough testing to back the operational claims, and that takes time.
FE: You advocate technology transfer and strong relationships. Were those manifested in this project?
Zargari: Yes. To achieve active front-end technology, we needed a power semiconductor. From 1998 to 2000, we worked with Mitsubishi to develop that device. Strong relations were also built with our university research partners.
Choke development was in collaboration with the power research team at Ryerson University. We eventually took what they developed and scaled it up to medium-voltage application. When you work with a university research program, the most important thing is to have a champion both in industry and at the university. Dr. Bin Wu filled the latter role. It’s also important to have regular meetings and to understand what has been done to keep both parties involved and engaged. Even when we moved to the industry prototype phase, we continued to keep our research partners updated.
FE: When did the university collaboration begin?
Zargari: We started working together around 1999. We had the idea earlier and had talked about how it could be done, but there was a lot of skepticism in industry. The challenge with a long-term project like this is that you have graduate students who want to publish and leave. Staying focused can be difficult. It’s very easy to go off target with the research.
FE: Once development moved in-house, what scale-up issues did you encounter?
Zargari: When we transferred the technology, it did what it was supposed to do but was not as cost- or size-efficient as we needed it to be. From an industrial point of view, that is a design consideration you can’t ignore. Originally there were tentative release dates in 2003, but we said, “We’re not there yet,” and took a big step back.
The prototype was a 120-V system with four windings on a core. It was the size of a telephone. When you scale up to 1,000 hp, the choke alone is 4 cubic feet. We had to change the design completely, and that added a year. We came up with ways to combine the windings and go from four to two. The commercial version looked entirely different than the prototype.
FE: Is elimination of the transformer the most significant advancement?
Zargari: Yes. We already had a choke in our topology, though you can’t take our choke and throw it into another motor drive and get the same results. Since 2004, no one has been able to duplicate it. That allowed us to quadruple our size in five years.
With AC drives, the waveform is converted to a DC wave, then back to AC with the desired amplitude and frequency. At a given voltage, the motor runs at a fixed speed. To change motor speed, variable frequency is required. There are different topologies to accomplish that, all with advantages and drawbacks. We believe our topology, with its inverter configuration for substantially sinusoidal output voltage and reduced switching losses, is a superior solution.
FE: The Innovator of the Year award comes with a $200,000 grant. How will you use it?
Zargari: Standards and consistency are the big issues in wind energy conversion systems. Utility companies don’t want power if it isn’t provided reliably. But if the topologies in AC drives were to work backwards, the variable speed of the wind could provide a constant feed to the power grid.
We’re working with the university on that kind of application. Building and testing a wind turbine takes a lot of money, and $200,000 won’t pay for more than a simulation study. Hopefully the economy will turn around and there will be some commercial opportunities for the technology, most likely in offshore wind farms.
