Sinovel Wind Stole wind turbine code from AMSC. |
After a three-hour trial, Dejan Karabasevic, 38, was sentenced to one year in jail and two years probation for passing proprietary turbine control software to Sinovel Wind Group Co., based in Beijing.
Karabasevic, who worked in American Superconductor’s subsidiary in Austria, was also ordered by Klagenfurt district court officials to pay his former employer roughly $270,000 in damages. “I made the biggest mistake of my life,’’ Karabasevic said before the jury began deliberations, “and I am deeply sorry for it.’’
American Superconductor is pursuing separate claims against Sinovel, which it believes is using the company’s electronic software code illegally. “This sentence was an important first legal step for AMSC,’’ said company spokesman Jason Fredette. “This individual’s admittance of guilt and collusion with Sinovel demonstrates the strength of our commercial, civil, and criminal cases against Sinovel in China.’’
American Superconductor, which makes control systems and other advanced technologies for wind turbines and power utilities, began to suspect Sinovel of theft in June, when an American Superconductor crew inspecting turbines in China found an imperfect replica of its software in a Sinovel turbine.
An investigation led American Superconductor officials to Karabasevic, who was arrested in Austria in July. Earlier this month, the company also filed suit against Sinovel, which denies any wrongdoing. Sinovel could not be reached for comment yesterday.
According to an indictment filed in the public prosecutor’s office in Klagenfurt, Karabasevic had developed “increasingly strong ties’’ with Sinovel as the “result of numerous business trips to China.’’ Karabasevic, the indictment states, was paid at least $20,700 to copy codes downloaded from American Superconductor’s network, and use them to help Sinovel upgrade versions of the US company’s software that he said the Chinese company had already “cracked.’’
Yesterday in court, Karabasevic’s lawyer, Günther Huainigg, said his client’s actions stemmed from “frustration’’ about a failed marriage, which had been strained by his trips abroad for work as a department manager involved in software development, followed by a demotion to the customer service department last year. Karabasevic believed he was undervalued at AMSC Windtec, the American Superconductor subsidiary in Klagenfurt, according to Huainigg.
Karabasevic in April downloaded American Superconductor software to his notebook computer and transferred it to Sinovel using his Gmail account, said Thomas Liensberger, the state prosecutor in Klagenfurt.
Florian Kremslehner, an attorney for American Superconductor, said there is evidence that Sinovel wooed Karabasevic by offering him an apartment, a five-year contract at twice his current pay, and “all the human contact’’ he wanted, “in particular, female co-workers.’’
Kremslehner said yesterday that American Superconductor now believes the theft of its software by Karabasevic was the catalyst that caused Sinovel - once its largest customer - to stop accepting shipments from the Devens company earlier this year.
As a result, American Superconductor’s stock plunged nearly 80 percent and it cut 150 jobs, about 30 percent of its workforce. The company yesterday reported a net loss of $37.7 million for its first quarter, and said revenues for that period were $9.1 million, down from $97.2 million a year ago.
In a conference call about its earnings yesterday, American Superconductor chief executive Daniel McGahn said Sinovel still owes the company at least $250 million for products it ordered. McGahn, however, remained upbeat, saying that he believes Karabasevic’s admission of guilt and conviction “demonstrates the strength of our case’’ against Sinovel.
The company’s stock rose 13.65 percent yesterday on the Nasdaq stock exchange, to close at $5.08 a share on word of Karabasevic’s conviction and the news that American Superconductor has signed nearly $100 million worth of new contracts since April with customers in China, India, South Korea, Europe, and the United States.
“While China has been the catalyst for our past growth,’’ McGahn said, “we expect our success, going forward, to be driven by a more diverse base of customers across a broad spectrum of geographies.’’
In a story that Hollywood would love to script, the U.S. Government charged Sinovel and its executives with soliciting the then-head of the Automation Engineering Department of AMSC’s Austrian subsidiary, AMSC Windtec GmbH, to steal AMSC’s source code in order that Sinovel might bypass a commercial relationship with AMSC and utilize AMSC’s trade secrets without paying for ongoing software licenses.
On June 27, 2013, a grand jury in the United States District Court for the Western District of Wisconsin charged Chinese-wind-turbine-manufacturer Sinovel Wind Group Co., Ltd. (“Sinovel”), two of its high-ranking executives, Su Liying and Zhao Haichun, and a former high-level AMSC engineer, Dejan Karabasevic, of conspiring to steal and actually stealing AMSC’s trade secrets, engaging in criminal copyright infringement, and engaging in wire fraud. The following facts are taken from the United States’ complaint and indictment.
AMSC, formerly known as American Superconductor, Inc., is a Delaware corporation, headquartered in Massachusetts. AMSC’s core business is the development, support, and production of equipment and software for wind turbines and electrical grids. Amongst its products is software utilized to regulate the flow of electricity from wind turbines to electrical grids. One issue over which wind turbine manufacturers have had to overcome in this emerging technology is how to maintain operation when a temporary sag or dip in the flow of electricity occurs over the electrical grid. Attempting to solve that issue, AMSC developed its Low Voltage Ride Through (LVRT) software, which it licensed to, among other companies, Sinoval. Sinovel sold and serviced wind turbines worldwide, including in China and the United States.
Until March 2011, Sinovel purchased software and equipment from AMSC. At that point, Sinovel owed AMSC over 100 million dollars and had contracted with AMSC to purchase another 700 million dollars’ worth of AMSC software, products, and services in the future. See, e.g., Daniel Cusick, Chinese wind power giant faces U.S. indictment on economic espionage charges, Environment & Energy Publishing (August 22, 2013, 9:23 a.m.)
In March 2011, without prior warning, Sinovel suddenly ceased accepting products from AMSC. Around the same time, specifically on March 10, 2011, Karabasevic submitted his resignation to AMSC, which was accepted on March 11, 2011. Karabasavic’s final day with AMSC was June 30, 2011.
In June 2011, while servicing wind turbines in China, AMSC engineers discovered that Sinovel was utilizing a version of AMSC’s LVRT software in a wind turbine in China, software which AMSC had not sold or licensed to Sinovel. Notably, fearing international espionage and unauthorized use of its software, AMSC previously developed an encryption code to prevent unauthorized use and a separate software code allowing only a two-week trial period of the software before the LVRT software stopped functioning. Inspecting the code more closely, AMSC determined that the code had been modified in two particular ways: (1) bypassing the encryption code by deactivating it; and (2) disabling the two-week test period software-use limitation.
Within the following weeks, AMSC investigated the activities of Karabasevic—in some instances utilizing a private investigative firm—discovering a host of email and Skype communications between Karabasevic and various Sinovel employees.
Those communications outlined a plan between Sinovel and Karabasevic whereby Karabasevic would download, provide, and, when necessary, modify and/or repair the AMSC LVRT software to work within Sinovel’s turbines.
For his work, Sinovel and Karabasevic agreed to a six-year, $1.7 million dollar contract, beginning in May 2011, which also allegedly included Sinovel providing Karabasevic with a Beijing apartment and sending funds to Karabasevic’s girlfriend in August 2011, a month before Karabasevic was convicted by an Austrian court of industrial espionage in September 2011.
Karabasevic was confronted by AMSC on June 30, 2011, and provided a signed statement on July 28, 2011. In his statement, Karabasevic admitted to the above. Karabasevic also gave written permission to search the Beijing apartment secured for him by Sinovel. During that search, two notebook computers, an external hard drive, and other documents were obtained. Subsequent review of those devices and materials corroborated that Karabasevic stored, possessed, and manipulated AMSC’s LVRT software.
Based on the FBI’s investigation, the FBI subsequently obtained a warrant to search various Sinovel wind turbines that Lumus Construction was scheduled to install in Massachusetts. In March 2012, May 2012, and December 2012, FBI agents searched such turbines and determined that AMSC’s pirated software was found within these Sinovel-manufactured turbines.
Since the United States obtained the indictments in late June—and in what bears particular attention for companies doing international business—Sinovel is attacking the efficacy of the United States’ summons, arguing that the United States’ service attempts are faulty insomuch as the United States improperly (and ineffectively) served Sinovel’s U.S.-based subsidiary, Sinovel Wind Group (USA) Co., Ltd., but not its Chinese parent, Sinovel Wind Group Co., Ltd. Sinovel argues that it is not subject to the United States’ summons power as a Chinese corporation and that service of Sinovel USA (dissolved in early July 2013) is not proper under Federal Rule of Civil Procedure 4.
On September 6, 2013, the United States filed its response in opposition to the motion to quash, arguing that Sinovel USA is the alter ego of the parent corporation. At the heart of this argument, the United States is claiming that Sinovel’s recent closure of its US subsidiary was made to avoid service of process and a direct result of the United States’ case against Sinovel.
This case—beyond providing an interesting alleged story of corporate espionage—should also present a bellwether, of sorts, for the United States’ prosecution of foreign corporations and may cause U.S. corporations to further evaluate their international business ventures with certain companies.
As noted in Sinovel’s moving papers, the United States has not had much, if any, success in properly serving foreign corporations in matters in which the United States seeks to bring criminal charges.
If the United States cannot hale foreign corporations accused of international espionage into its courts, then—no matter the investigative lengths to which the United States goes—United States corporations may continue to lack adequate protections against foreign corporations accused of stealing their trade secrets. This is particularly the case whereas here the indicted individuals are all foreign nationals whom the United States has little, to no chance, to ever extradite and is further exacerbated by AMSC’s repeated losses against Sinovel for its business losses before Chinese Courts.
While failure to prosecute Sinovel is a worse-case scenario for the United States, this case continues to highlight why companies must monitor the access its employees have to secure databases and to a company’s trade secret and confidential information.
Companies should consider using additional preventive means to prohibit employees from stealing trade secrets, such as configuring the operating system to restrict access to external devices, thus, restricting the ability to download information to an external device; blocking a user from uploading information to a web-based site; and/or utilizing software that blocks employees from sending emails to certain domain names.
In situations like this, companies may also wish to consider placing blocks on the ability of its employees to email certain domain names that are known to be used for personal email accounts. In an era in which data is becoming increasingly portable, companies much increase their vigilance in monitoring use and exporting of its data and trade secrets.
A company living off licensing its wind turbine design to manufacturers – is that a viable business plan? Timothy D. Poor explains AMSC’s business model to Renewable Energy Focus U.S.’s Kari Williamson.
American Superconductor (AMSC) has been largely successful with its Windtec subsidiary, through which it licenses out wind turbine design – from foundations to blades – but how can that be a profitable business model? Senior Vice President, Global Sales and Business Development, Timothy D. Poor, explains: “You can think of our Windtec business just like a ‘Vestas’ or a ‘GE’, except we don’t have a wind turbine manufacturing plant.”
“We sell wind turbine designs including designs for the manufacturing plant to our licensees. We help with the supply chain, which is very important. We do everything that a wind turbine manufacturer would do, except we don’t actually make the turbines – it lets us focus on where we add the most value, and we add value on the engineering side, and on the design and the system level optimization of the wind turbine.” AMSC also helps its customers with sales and marketing of their systems.
AMSC Windtec in Short
AMSC Windtec provides licenses and customized designs for onshore and offshore turbines ranging in power from 600 kW to 10 MW. Licensees receive unlimited service and support from the prototype development stage to full-scale scale production. AMSC Windtec additionally helps its customers to establish a local supply chain.
AMSC Windtec also develops complete electrical systems for wind turbine applications. Based on the company’s PowerModule™ power converter, these electrical systems manage power flows, control the operation of the entire wind turbine and monitor the turbine’s performance with a S CADA and Condition Monitoring S ystem.
AMSC’s wind business has grown from approximately US$13 million to over US$250m in sales over the last three years.
But is licensing revenue enough to make a profit? “How we make our money is actually selling the electrical components that go into the wind turbine. We sell you the license and transfer he technology and become part of the supply chain for you,” Poor explains.
“So all the electrical controls that go into the wind turbine, the power converter – the brains of the wind turbine if you will – are coming from us, and we sell those to our licensees on an ongoing basis.”
Another part of the AMSC Windtec business model that, according to Poor, gives AMSC an advantage, is that they are not committed to a particular drive train type or wind turbine design. “We can switch,” he says. “In every AMSC case, our only mission is to build the best-performing wind turbine. It doesn’t matter which drivetrain or technology we use, or where they’re coming from. We think that’s a unique starting point for a company.”
Rated vs. Actual Output
Up in the nacelle, Poor says there is a trend to use permanent magnet type generators using a full-scale converter. AMSC licensees Hyundai and Doosan both use full conversion in their multi-MW wind turbines.
“At partial loading, a permanent magnet generator is more efficient– and has less losses. … Wind turbines typically only run at a fraction of their top power rating, so the efficiency gains on a permanent magnet machine – especially for your larger megawatt sizes – are not insignificant. Those factors coupled with the support to the grid enabled by the full scale converter make them very attractive designs for wind turbines.
“There’s also a trend to larger rotor diameters, so bigger blades. The whole game with wind is ‘how close can you run to that rated power setting?’,” Poor says. “There is a focus by all of the global players now towards making bigger and stronger and lighter blades.”
Although the race towards higher power ratings is still on, it is no longer the ‘only goal’ of the wind turbine industry. Increasingly, the trend is to get as close to the capacity rating as possible – optimizing the output of existing platforms.
“We’re working on a 100 m rotor for a 2 MW machine. These are focused at sites that have slower wind speeds.” Poor explains: “Every little bit you can increase [actual output] is money in the pocket of the developer. Putting bigger rotors on turbines opens up the markets for certain sites that could not be economically justified to be developed before – suddenly now they become economic. It will increase the size of the market, quite frankly, if manufacturers were able to do that.”
Blade Dynamics
Blade Dynamic’s technologies can enable blades longer than 90 m for the next generation of turbines, such as the seaTitan. The rotor is lighter, more powerful and more cost effective than most existing technology. Blade Dynamics says the rotor increases annual energy production (AEP) by as much as 10%.
The rotor manufacturer has also developed a surface coating, Bladeskyn, said to last longer than traditional paint and gelcoat. It features UV stability, minimal dirt pickup, abrasion resistance and very low reflectivity.
AMSC additionally recently invested in a UK company called Blade Dynamics on the Isle of Wight, which designs and manufactures wind turbine blade technologies engineered to increase the efficiency and performance of very high power (multi-megawatt) wind turbines while also reducing costs.
AMSC now owns 25% of the company and helps promote Blade Dynamics’ technology also to non-AMSC licensees. “The neat thing about Blade Dynamics’ technology is that it’s 40-50% lighter than conventional blades. That means we can put very large rotors on turbines and open up markets that aren’t economical otherwise.”
This especially applies to low wind sites: “With a very big rotor integrated efficiently into the design of our wind turbines – we have a very good product for that market. We think the US and North America are prime markets for this, as well as Europe, and Asia.” Blade Dynamics is, amongst other things, working on the blades for AMSC’s 10 MW SeaTitan offshore wind turbine.
The SeaTitan
A few companies, including AMSC, are currently working on 10 MW wind turbine designs. The most prominent being Clipper’s Britannia project; Sway of Norway; and AMSC’s SeaTitan, which utilizes superconductor direct drive generators that are equivalent in size and weight to a 5 MW system utilizing conventional technologies.
“Nobody’s building anything yet, so it’s going to be interesting to see how this works. Our focus is not just to build a 10 MW wind turbine – anyone could do this. But you want to build one that’s commercially viable and that’s going to be economic for the development of large-scale offshore wind farms. You’ve got to take into account all of the costs and logistics associated with building an offshore farm including the subsea infrastructure.
“The SeaTitan is much smaller in size and lighter in weight at the top of the tower, which enables much lower installed costs for an offshore wind farm. About 60-70%, or two-thirds, of the cost of an offshore wind farm is not in the actual wind turbine itself, but in the foundation – the construction of the foundation and the installation of it into the sea.
10 MW SeaTitan Offshore Turbine
AMSC’s 10 MW SeaTitan offshore wind turbine, which is expected to enter production around 2015, utilizes high temperature superconductor (HTS) rotors rather than copper in the generator. This enables the generator to be much smaller, lighter, more efficient and less expensive than conventional large-scale wind turbine generators, according to AMSC Windtec.
SeaTitan will use a direct drive generator, which could lead to less maintenance requirements than a gearbox. The wind turbine is expected to have a rotor diameter of 190 m, and a tower height of 125 m.
“If you can keep those costs somewhat fixed and yet raise dramatically the power rating of the wind turbine, you’ve definitely improved the economics of the wind farm. … We’re taking the infrastructure you need for a 5 MW turbine and we’re putting 10 MW at the top of the tower. So you’ve got essentially the same costs as you would in terms of all the infrastructure, but for 5 MW you get 10 MW.”
Asian Power
AMSC says its business model serves it very well in Asia, and particularly in China where it currently has a total of five customers. When China decided it would have a domestic wind industry a few years ago, it faced the problem of not having the technology. AMSC as a technology provider could therefore enter the Chinese market without worrying about the Chinese requirement to set up shop in the country.
Sinovel Wind is one of AMSC Windtec’s China-based customers, and according to Poor, it is the third largest wind turbine manufacturer in the world and the largest in China. “They’re using all our technology, buying all their components from us on an ongoing basis.” Dongfang is another large customer.
Looking West
“Going forward, Asia – and China in particular – will continue to be the most important market for us in terms of where our revenue is coming from, but we’re also focused now more on the West. We eventually expect to have additional licensees in America and Europe – we’re actively now out promoting this business model to Western companies,” Poor says, before adding: “We don’t feel there’s any reason why we can’t be just as successful in the West as we have been so far in Asia.”
AMSC Windtec believes its technology will be part of major offshore wind farm developments such as the ones under the UK Round 3. “We have three licensees making 5 MW turbines, which is ideal for offshore – Dongfang, Sinovel and Hyundai Heavy Industries. Over the next year, they will be erecting and commissioning their first 5 MW turbines. They will be ready to go into commercial production after that, and we expect them to benefit from the Round 3.”
As Europe could be seen as the ‘headquarters’ for wind power, Poor believes it is a very competitive market. AMSC Windtec is therefore concentrating on the less developed offshore wind market. “From the European perspective, that’s a segment that interests us most and where we see a lot of potential.”
US Market Potential
The US and North American market as a whole, represents a vast market for the company based in Devens, MA. “We’re focused on – for markets onshore like North America – really high performance turbines,” Poor says.
It looks like AMSC Windtec is heading
back to its origins – North America – with its Windec business model. The
question is, who will take up the baton to become the Sinovels and Dongfangs of
North America?