News and New Products

Is there gold in Green Technology? A microelectronics perspective

A Silicon Valley panel mines for the realities behind the Green Tech hype.

By Ron Wilson, Executive Editor -- EDN, 9/18/2009

A panel this week sponsored by international science and technology interest group Monte Jade explored the realities behind the promise of huge new microelectronics markets from so-called Green Tech—the global drive to lower carbon footprints, reduce energy use, and shift to sustainable consumption. The panel combined venture capitalists, industry executives, and an executive from a key electric power industry organization, and produced a fascinating cross-fire of ideas.

The first question posed to the panel was a reality check: is there really an opportunity for microelectronics vendors in the Green Tech movement, and if so, where? First to respond was Jim Diller, a angel-level fund leader at venture capital group VentureTech Alliance. VentureTech has holdings in, among other things, LED light-source start-ups. Diller pointed out that today about half of the world is underserved in terms of lighting. Given the huge advantage of LEDs in efficiency and life, he said, a huge market for LEDs is a question of when, not if. He said that other markets, such as solar energy, would also need microelectronics for measuring and monitoring functions. But he added that these markets are not made up of electronics experts, and need solutions, not just parts.

Bill Orner, VP engineering at lighting engine vendor OptoElectronix, agreed that there was a huge potential lighting market. But he emphasized that this was going to be a different market than the industry infrastructure was used to attacking. “The value here is in little microcontrollers, not in server CPUs,” he said.

Erfan Ibrahim, Technical Executive, AMI/Han Lead, Power Delivery & Utilization, at the Electric Power Research Institute came at the answer from a different perspective. “I believe that the next 30 years in the US will see the largest-ever growth in the consumption of electronics,” he said. “The impact from the build-out of the Smart Electric Power Grid will be huge.”

Sudhir Mallya, director of SoC solutions at NEC Electronics America, spelled out three concrete opportunities in Green Tech: low-power semiconductors to reduce power consumption in end-products; intelligence in appliances and machines, also to reduce their energy consumption; and, agreeing with Ibrahim, the Smart Grid, which will require both local intelligence and communications technology.

But Chih Kai Cheng, Founder and General Partner, Harbinger Ventures, emphasized the complexity and challenge of these new markets. “There is a range of semiconductor applications within Green Tech,” Cheng said. “This could be a huge opportunity, but it is very fragmented, and the amount of knowledge necessary to make good choices in this area is very great.” As examples of the complexity of the markets, Cheng said that there were still opportunities in solar cells. But there are at least two competing technical approaches, and for each of them conversion efficiency, cost, and market size will all interact.

The panel then dug deeper into the question of what would be necessary to make a useful return on investment on these opportunities. Cheng’s response was that everything hinges on the behavior of governments. He suggested that we could learn from the European Union experience. The EU chose to emphasize consumption taxes instead of efficiency incentives, Cheng said. In fact, Cheng said, the EU expects to derive 50 percent of its tax revenue from consumption taxes in ten years. The impact of this choice is that EU per-capita energy consumption is one-third that of US cities.

Mallya warned that Green Tech would not be business as usual for the semiconductor industry. “The opportunity is very different from previous booms we’ve seen,” he said. “This time, the semiconductor industry can’t push a solution onto the market and expect people to adopt it. And the role of government will be critical.”

Ibrahim also emphasized that the industry would have to think differently to have a role in the new market. They would have to get beyond the distinction between themselves and their customers. “There is a saying by the Persian Poet Rumi,” Ibrahim quoted. “ Out beyond ideas of wrongdoing and rightdoing, there is a field. I’ll meet you there.”

Explaining, Ibrahim offered input he has heard from electric utilities that participate in the Research Institute. First, he said that the microelectronics industry had to understand that the Smart Grid has been declared a critical infrastructure, placing it under the eyes of the Department of Homeland Security and other agencies. One of the implications of this is standards: utilities absolutely will not accept proprietary solutions that would bind them to a single company, or that might have undiscovered vulnerabilities. Another point, he said, was “Don’t design the chip first, and then go looking for a customer. Go out and understand the use cases: the utilities have been developing them, and they are available. Then when you understand the use cases, develop a solution.” Finally, he cautioned vendors to “think solutions, not devices.”

Orner agreed with Ibrahim’s warning. “The things that made us so successful in the consumer electronics market are exactly the wrong approaches for Green Tech,” he said. “Number one among these is product churn. In consumer, you start a new chip design as you are taping out the old one. But in many of these new areas, the customers need a ten-year or longer product life to get return on their designs. They won’t accept our short-term thinking.”

Orner went on to add that in his view, government support, with its political unpredictability, was a mixed blessing. Rather than stalking government incentive money, Orner suggested, the most reliable opportunities came when you could show the end customer either an attractive payback period, a competitive advantage, or a best route to regulatory compliance.

But Diller countered that the government was a major factor in opening Green markets. “Look at Title 24 in California,” he said. “Since that measure was adopted, electric power use has been flat in the state, while it has grown by a factor of two in the rest of the country.” Diller also cautioned, though, that Green Tech markets were about solutions to problems, not about merchandizing. “We tend to get wrapped up in our own technology,” he admitted. “Get over it.”

Asked by an audience member to define the Smart Grid, Ibrahim delivered a remarkably concise accounting. “First,” he said, “think of the Smart Grid as the bridge between where we are today and a future in which wide-scale nuclear generation and carbon sequestration are both practical technologies. Smart Grid is a way to defer the growth in energy demand until we have the generating capacity to meet it.”

Ibrahim then defined what the Smart Grid was not. “Despite what a number of semiconductor and networking companies think, this is not about slapping a control network on top of a 19th-century electric power grid. In fact, power equipment makers have been looking ahead and designing intelligence into their equipment for years. The intelligence is there. What is not there is the public network co-located with the equipment to connect it together.”

Then Ibrahim mapped out the Smart Grid in three sets of triplets. Technologically, he said, the Smart part was based on three major components: smart meters, Phasor Measurement Units (PMUs), and home energy controllers. Smart meters, already familiar to many people, are now primarily used for remote reading to cut down the cost of billing. But they have the capability to be much more, including both state monitors at the edge of the grid to allow the utility to measure actual demand, and consumption indicators for the individual customer.

PMUs, in contrast, are instruments that can make synchronized instantaneous measurements of magnitude and phase at many points around the grid. Many theorists of electrical-grid dynamic behavior feel PMUs are the most important sensors for estimating the state of the grid and controlling it. Ibrahim said that today, PMUs in a grid typically make synchronized measurements a few times a second. This needs to be moved up to about 60 samples/second to enable modern control algorithms.

Finally, Ibrahim describe an energy-management service box: a programmable, policy-based controller that would reside inside a subscriber’s home, collect data from the smart meter on the outside wall, the appliances inside the house, and the controls and switches operated by the residents. In this scenario you wouldn’t turn on a light or set a thermostat so much as you would send a request to the service box, which would then use the current state of the network, your current consumption, and the estimated energy load you were requesting or releasing to comply with your request, inform you about the cost of the energy you were about to use, or simply override your request in extreme situations.

Ibrahim said that these technologies came together to meet three strategic objectives, each of which the Electric Power Research Institute studies: awareness of the grid state—what is called wide-area situational awareness; energy storage in the grid; and energy transportation through the grid.

Innovations will not be rapid, Ibrahim warned. “Our policy is crawl, walk, run. First we investigate a technology to understand its feasibility, then we do larger-scale tests in an isolated portion of the real grid to make sure it works, and then we deploy it.” Ibrahim said that the process was somewhat easier than FDA approval for a new medical device, but far from fast. “If you need quick return on investment in Smart Grid technology, go for a technology that is already approved,” he counseled.

Asked for a realistic assessment of the progress in the Green Tech world, the panelists were cautious. “So far, the report cards for Smart Grid tests have been all over the place,” Ibrahim said. “Some utilities have reported breaking even on their investment just by using remote sensing with smart meters. Some have seen returns on optimizing power delivery strategies.”

But Cheng warned that breaking even was more than a financial matter—it had to consider the entire cost of a new development. “Think about the refrigerator in your kitchen,” he suggested. “Right now it can’t work with a service box, because it has no local intelligence. You could buy a new smart refrigerator. But if you did, would you ever recover the cost to the environment of building it?” In an aside, the panelists generally agreed that savings from putting more efficient cars on the road with the US cash for clunkers program would never recover the environmental cost of building the new cars.

And there are people issues as well. Semiconductor engineers, electrical-power engineers, and lighting designers don’t communicate well with each other. Hasty marketing sometimes creates disasters, as in the bungled launch of compact florescent lamps. “I hope the LED industry doesn’t screw it up like the CFL business did,” Diller worried.

And the stakes are large. About half of the labor force in the electric utilities in the US will retire soon,” Ibrahim warned. “Their average age is 48.” He said that to deal with the decline in experienced field technicians, the industry needed a self-diagnosing grid, asset-management systems to tell technicians not only where the failure is but what’s installed out there, and smart dispatch to send the right technician and equipment to the right place. And we need to reduce the growth in demand, fast.

“In the next five to ten years we will have brown-outs, as the peak load grows more quickly than the peak capacity on the grid,” he said. “In 15 to 20 years, generating capacity in absolute terms will be less than demand.

“This is not a convenience issue,” Ibrahim concluded. “Historically, strategic advantage in the world has not been about armies or weapons. It has been about the ability to provide and control energy. This is not a problem we can set aside.”

And that perhaps offered the best answer to the original question. There are many difficulties in finding profitable semiconductor markets within the Green Tech wave. But the alternative to ensuring that Green Tech succeeds is to see the industrialized nations gradually subside into developing economies, with nowhere left to develop.