At the MIT Clean Energy Prize, held May 11th in Cambridge, Massachusetts, an entrepreneurial team looking to disrupt copper’s dominance in electricity transmission took home both the Department of Energy’s $75,000 award and the MIT Grand Prize of $200,000. OptiBit, a start-up composed of young scientists from CU Boulder, UC Berkeley, and MIT, is looking to harvest massive energy savings, principally targeting the burgeoning cloud, where the world’s data continues to run along copper wires, but which, OptiBit’s Chen Sun remarked “could be replaced by light.” OptiBit’s fibre optic innovation was clearly part of a trend at this year’s MIT award show and competition.
Indeed, “clean tech” no longer seems the right term for today’s energy startups. “Efficiency tech” or “booster tech might be more appropriate. It is as though the current field of energy entrepreneurs have correctly intuited that clean power generation, already coming on strong, is ready to fly under its own momentum. What’s needed now is the technology to route, enhance, conserve, and maximize the electricity that’s increasingly coming from renewables.
During the Talking Points Memo five-part series on renewables, we’ve examined the extraordinary growth in global solar capacity, which should reach 500 GW by the end of the decade, compared to 175 GW today. We’ve also charted the stagnation of global coal growth, as aging plants retire in the US and are also increasingly shut down in China.. Global wind power meanwhile, which got a head start on solar early in the last decade, currently stands at roughly 370 GW of capacity. In fact, combined wind and solar together are growing so fast that in 2013, they provided over 12% of new energy supply from all sources.
While it’s not yet clear how existing fossil fuel consumption globally will be meaningfully pulled lower just yet, some of the excitement that surrounded Tesla’s recent home battery unveiling suggest many now think the pathway to a fossil-fuel free future is, while still fuzzy, finally coming into view. Because storage will play such a key role in a future smart grid, it’s understandable the Tesla’s 10 kW Powerwall battery shocked the market with its unexpectedly low starting price of just $3500. Sales of the home storage unit sold out within the first two weeks, generating $800 million in reservations. That caused one commentator to write, “We now have just about everything we need for a technological fix for climate change.” Really? Is that true now, or will it be true in the future? Let’s imagine a hypothetical scenario.
It’s late September in the year 2021, and Dave Fisher of San Bernardino, California is ecstatic. Equipped with a 200 kW, grid-connected home battery array from Tesla, Dave has just heard the Santa Ana winds are about to blow. Using a Google app on his iPhone that controls his home’s smart-thermostat, Nest, he enters a buy order to download cheap, surplus wind power he expects to come off California’s grid overnight. The Santa Ana winds, blowing south across the Mojave at a constant 30 mph, will also completely burn off Southern California’s smog the following day. So Dave enters another buy order for even cheaper surplus solar power.
By dinnertime the following evening, Dave can’t believe his luck. His bid to buy excess wind power was filled at 9 cents kilowatt, and his order to capture the expected solar overgeneration was filled at 7 cents a kilowatt. Indeed, there was so much solar surplus on California’s grid in the afternoon, he charged up his EV’s battery as well. And now it was time for the final coup de grace. Dave calls up his friend Barry and says, “Are we ready for the trifecta? Watch me sell some of this back to the grid!” Dave enters a sell order to discharge half of his battery for 22 cents a kilowatt. As Californians come home from work, demand spikes massively across the LA Basin, and his offer is hit immediately. “I can’t believe how you are gaming this system, “ Barry says wearily. “Hey, you should thank me,” says Dave. “I’m providing liquidity, and by doing this every day I’m helping to shift loads in our age of renewables.”
If energy transition is to mean anything it should mark the historic shift to a world of energy capture, from a world of energy combustion. Such a transformation will make a lot of new-new things possible. In energy combustion, heat is wasted, and it’s harder to recover. In electricity transmission, however, power is wasted also–but it’s a much smaller amount and a tech-driven grid, yes, a smart grid, would be able to claw back transmission losses more effectively.
This is why despite the high infrastructure cost of energy transition, most models of a world running primarily on renewables forecast not just large, but gargantuan gains from efficiency. A more liquid power grid, a more responsive power grid, will enable individuals and devices to interact in such a way as to capture those gains. And those gains will come primarily from the technology that connects clean power generation. To capture the prize of efficiency, an internet of things is perhaps the best analogy for the power system yet to come.
“Well, I don’t think most people are interested in becoming energy traders. But in the economics of home-energy arbitrage, the likely scenario is a third party will eventually do that buying and selling for you,” says Constantine Samaras, an Assistant Professor of Civil and Environmental Engineering, at Carnegie Mellon University in Pittsburgh, Pennsylvania.
Samaras, who was a lead contributing author to the 2012 Global Energy Assessment and publishes frequently on energy transition, renewables, and electric vehicles, does agree that some sort of mesh, a kind of intelligent layer, will need to be incorporated into the future power grid. “Let’s say I have a solar-enabled home, and an internet-of-things array of appliances, and a home battery—when all these assets are eventually aggregated, then they can truly be put into play. But I’d rather that some entity tied them all together for me.”
The Energy Information Administration, in a recent update this April, showed that deployment of two-way smart meters has been moving so quickly that in 2013 they began to outnumber one-way smart-metering devices. “Two-way or AMI (advanced metering infrastructure) meters allow utilities and customers to interact to support smart consumption applications. By the end of 2013, there were 51.9 million meters operating in AMI mode…” reported the EIA.
Smart meter growth, dominated mostly by Texas and western states (California especially) grew by over 17% in a single year. While impressive, however, these kinds of deployments are surely just the leading edge of a long, unfolding wave of device upgrades that will be required to make the grid responsive. “The way I think about this—and there’s an analogy to the development path of telecommunications—is that in both distributed systems and smart infrastructure in the home, we are going to evolve from a few-to-many to a many-to-many network,” says Samaras.
In a 2014 study led by Mark Z. Jacobson of Stanford University, a future California was modeled in which every aspect of the state’s energy usage—heating, cooling, power and, yes, even transport—would be running on 100% renewables. The scale of the requirements is sobering. Of the 625 GW of renewable capacity modeled, over 425 GW would be composed of solar, with utility scale solar installations taking about 300 GW of that share, and rooftop solar accounting for the remainder, at 125 GW. Wind requirements would reach 170 GW of capacity, split between 130 GW of onshore and nearly 40 GW of offshore installations. The remaining 30 GW in the Jacobson model would be composed of tidal, geothermal, and hydro. By comparison, California today has roughly 6 GW of utility scale solar operating, with myriad new projects continually underway. To create a 100% renewable California would be a massive infrastructure undertaking, to say the least.
The study, “A roadmap for repowering California for all purposes with wind, water, and sunlight” is full of several, intriguing quantifications. For example, utility-grade solar is projected to take up less than 1% of California’s total land area. With vast stretches of emptiness in Riverside and San Bernardino counties alone, that would not present a barrier.
Rooftop solar is projected to take up even less space: just a quarter of one percent. Given the sprawling built environment already in place in the Golden State, that too should not be a problem. Where the modeling starts to get more aggressive, however, is in the projected savings from the transition from fossil fuels to electricity. Indeed, the model projects that by the year 2050, so much end-use efficiency is achieved, combined with the net energy reduction from transitioning from combustion, that total California energy demand is free to grow, even as its energy system is decarbonized.
Whether one believes such a transition is technically, or politically feasible, it’s crucial however to recognize that a well-established phenomenon known as the network effect would certainly be in play in such a feat of engineering, and in the type of energy system a modern economy could—at least theoretically—construct.
The Jacobson transition model, along with other transition models over the past few years, therefore posits not only the thermodynamic savings from switching to clean power generation, but also, the efficiencies that would come from network effects. While many of these models are less detailed on cost, those too are typically offset by considering inclusive, whole system gains not only from energy savings but from a rapid decline in health care costs, and other costly externalities associated with coal, gas, and oil combustion.
Consider, by contrast, how our system operates today. Transportation still runs mainly on the internal combustion engine—a powerful machine no doubt but one which expends (wastes) an enormous quantity of heat. Meanwhile, despite the advent of supercritical, or yes ultra-supercritical power plants burning coal or natural gas, heat loss in power generation is still non-trivial. Again, in Part III of this series which focused on Los Angeles, despite the heroic build out of electrified rail transport, and the halt in growth of the automobile complex, at least 75%-90% (it’s hard to quantify) of the total energy consumed in LA County still comes from combustion. And at some point along the cost curve, that type of system begins to throw off a increasing array of losses. Losses to health, and eventually losses to wealth. Yes, at some point—and this is what the Jacobson study and other studies in environmental economics suggest—your economic return from burning fossil fuels goes into decline. Perhaps even, rapid decline.
It’s easy to dismiss the prospects for an all-renewable energy system. As Jigar Shah, founder of SunEdison, quipped in his keynote speech at the MIT Clean Energy Price awards, “To accomplish big things on the scale of the systems we’re running today… is pretty much always going to cost, you know, a trillion dollars.” And that’s exactly right, because the value of the infrastructure already deployed today runs into the many trillions of dollars. The Jacobson study, for example, projects a $1.1 trillion capital cost to repower California on wind, water, and sunlight. (As the paper calls it: WWS).
So while these numbers seem foreboding, the work actually proceeds by one step at a time. Consider, for example, just one company—Solar City—which intends to install a full GW of solar capacity this year, while at the same time building manufacturing capacity to produce a full GW of solar panels each year. Not content to stop there, Solar City has predictably partnered with Tesla to smooth the installation of home-storage batteries. You could almost say Solar City is becoming a utility. A small utility today, but perhaps sooner than many expect, a very large, very dominant utility.
While competition is too fierce in emerging industries to make long-term predictions, it does appear that Solar City is an example of an emerging platform company in the renewable space. It is likely through platforms such as these (there will need to be many more) that the integrated power grid, responsive to supply and demand, will come into being. Carnegie Mellon’s Constantine Samaras remarks that while the new energy generation is of course critical, it will be the data coursing through new platforms that may drive each new iteration of the powergrid. “ I actually think the energy itself will become secondary to the information. So it could even be that energy will eventually become the dividend, because we want to build out this intelligent grid anyway.”
If the future power grid is going to become as responsive and as networked, say, as the internet, and energy information will be a currency as critical as energy flows, then it’s hard to see how Silicon Valley doesn’t get in the game. Companies like Uber are really just early routers of energy machines, harvesting efficiencies from existing rolling stock, and coordinating energy consumption through algorithmic signaling.
The primary task of a future smart grid will of course center on what economists call intertemporal time-shifting: harvesting the peaks of wind and solar generation, laying those surpluses onto networked storage, and then creating incentives for users to sync up their behavior with those enormous oscillations. Today’s giants of information technology like Google, device makers like Apple, and cloud service operators like Amazon Web Services would already be in position, it would seem, to leverage their platforms for such operations. But one development will be critical: enough users have to join in the network to create economies of scale, triggering network effects.
As if to underscore MIT’s expanding role in energy science and innovation, the Cambridge-based institution also launched in May a Future of Solar Energy study, presented in Washington, DC on May 5th. Of particular interest was the talk given by Vladimir Bulović, who explained that while silicon-based solar PV manufacturing is well-established today, and has great visibility to scale up further, it behooves material scientists to continue research on thin film solar technologies. And here’s why: while cost reductions have been extraordinary in silicon based solar so far, it’s now possible to see how the array of other costs around its deployment may slow those cost declines in the future. To diversify the solar mix, it would be ideal to have other competitive technologies. Alas, the problem with thin film, despite its promise, is the scarcity of needed rare elements.
The typical problems in energy transition are like a volatile sequence of promising cost declines, technical roadblocks, and shifting behavior that smoothes out bumpy pathways, all overwhelmed by the frustrations of time. Right as solar and wind deployments seem to be going much faster than expected, one is reminded that the global industrial economy is still on course, using the climate measure, to well exceed 2C of warming. Conversely, just as it seemed the Non-OECD would begin another round of coal adoption, it’s now clear that China and perhaps even India—leveraging their economic momentum—will divert enormous tranches of new energy growth into renewables.
If there is a sense of impatience that’s now associated with our energy transition, Carnegie Mellon’s Samaras’ says there may be a silver lining to this particular kind of forward motion. “I think about this problem a lot. Designing for long-term infrastructure under the conditions of uncertainty. We try to march up the curve of minimum costs and maximum benefits, without regretting too many of our choices. But even if we had the ability to convert all the infrastructure, right now, we need to try to avoid getting locked in, and being too unprepared for new disruptions. We should march up innovation and cost-reduction curves as we go along steadily. This is something that will take some time, but that pace may be a benefit.”
As this five-part series on renewables now comes to a close, it’s worth point out how much the conversation surrounding renewables has changed, in just the past three years. Many energy experts, even those constructive on wind and solar power, have long been skeptical the requisite adoption rates and market solutions would appear so soon. Now, even the most vocal doubters have watched the prospect of energy transition flip from a long, tough slog into a visible trend that’s now accelerating. From California to Africa, throwing cheap solar panels on rooftop is now easier, simpler, and faster than building a new, high-tech fossil fuel burning powerplant. Individually those panels are small. But collectively they are not merely providing new generation of clean power. More importantly, they represent the driver for a suite of energy technology solutions that everyone from Tesla, to European industrial giants like Siemens, is now furiously chasing.
To the point of Constantine Samaras, this is all unfolding at the point where uncertainty keeps dissolving into something like visibility. And if you look hard enough, you can begin to see how the world’s energy system is now in position to look very different just ten years from now. Whether that will be enough to allay the projections of rising temperatures and sea levels is not known. But the prospect of a radically-different energy landscape within the lifetimes of most people alive today is now appearing, on the horizon.