In politics, like in business, positive actions often have unintended, negative side effects. New research from Serguei Netessine, a Wharton professor of operations, information and decisions, reveals that very paradox in an economic policy long regarded as sensible and sound. The practice of taxing traditional fossil fuels is designed to promote renewable energies, such as wind and solar, which don’t pollute the environment with harmful carbon emissions. But the so-called carbon tax may not be the best solution for a greener planet. Netessine talked to Knowledge at Wharton about why creating industrial-scale ways to store renewable energy is a better way to go. The research was co-authored with Sam Aflaki, a professor at HEC Paris.
An edited transcript of the conversation follows.
Knowledge at Wharton: Could you give us a brief overview of your paper?
Serguei Netessine: The idea of the paper is really very simple. When you look at all kinds of solutions or approaches that are offered to increase the generation of electricity through renewable energy sources — like solar, wind and so on — one of the solutions offered is taxing carbon.
Knowledge at Wharton: If you tax fossil fuels more, you’re likely to encourage more renewables, correct?
Netessine: That’s what you would think and what the conventional wisdom says, but this is not quite what we found in the paper. There is a subtle, subtle reason behind it. When you try to rely extensively on renewable energy sources, the problem is that pretty much all renewable energy sources are intermittent. That is to say, the sun doesn’t shine at night and if you are to rely exclusively on solar energy, you will be in a lot of trouble because most of the consumption of electricity happens just before the sun comes up or just after the sun goes down.
What do you do? In this case, there is a grid operator who has to balance a grid. The grid operator would love to use renewable energy because the marginal cost of it is zero. The sun doesn’t really cost you anything. When you want to balance supply with demand, you have to bring online some kind of an energy source that can be ramped up and ramped down very quickly. Unfortunately, the only reasonable way to do it is to use heavily emitting energy sources like gas-fired turbines or coal. Gas turbines are very flexible. You can ramp them up, you can ramp them down. But pretty much for every megawatt hour that you would generate using solar, you need a backup. Since you need this backup, you have to have it somewhere in reserve, and sometimes you’ll bring it online and sometimes you power it down. When you use it like that, it needs a lot of carbon. So, the more solar you’re going to use, the more backup you’re going to use. If you’re going to tax carbon, you might actually end up having more of those backup energy sources that are heavily emitting [greenhouse gases]. That’s the gist of the paper.
Knowledge at Wharton: What are the alternatives to taxing carbon as a way to promote renewables?
Netessine: There are a couple of things we can do. One thing you can do is deal with intermittency. Of course, there’s not much you can do with the sun and the wind — you cannot change that. But think about storage technologies. Currently, storage technologies are nowhere at the industrial scale. There are lots of prototypes and ideas. Some people are talking about, for example, pumping water up the hill and then letting it flow down when you need electricity. Or you can push the train up the hill and then let it roll. But the most realistic one we have right now is battery-based storage, and Tesla is trying to bring it to the industrial scale, although we are nowhere close yet. That’s the whole problem because if you could store this electrical energy during the day and use it during the night, then in combination with solar technology you would be fine.
Knowledge at Wharton: When you say less intermittent, what does that mean?
Netessine: Basically, I’m going to pay you more for electricity if you can give me stable electricity…. Electricity generators don’t really worry about intermittency. They say, “When I have electricity, I just give it to the grid and the operator pays me.” But if you incentivize them to be less intermittent, they would start thinking about pooling different sources of electricity together. Let’s say I have a hill and wind might be blowing on one side. If I don’t have incentive, I just build turbines on one side…. But if I were incentivized to be less intermittent, I would maybe build on both sides. Or maybe I would build solar and wind, which are not that correlated, so I would be less intermittent. Currently, this is not happening.
Knowledge at Wharton: There’s the idea of a household using its electric car or maybe some other battery to store energy during the day and use it at night. Do those batteries have enough capacity to do that?
Netessine: No, this is not industrial scale right now. More importantly, you have to ask a question of producing batteries. It’s a very dirty process environmentally, so it’s not entirely clear to me that this would be an environmentally friendly solution. The more immediate solution would be this kind of a pooling of capacity and using different renewable energy sources together to reduce this intermittency. I think that is a more short-term solution.
Knowledge at Wharton: You reduce the intermittency, but you’re still going to require some kind of fossil fuel to help with those periods where you can’t quite cover what’s going on.
Netessine: Absolutely, and that’s the problem with balancing an electrical grid. You cannot have a situation where demand exceeds supply, otherwise you end up in a blackout. The grid operator has to work really hard, otherwise you have to start selectively shutting down the people who consume a lot of electricity, and nobody wants that.
“There are many arguments against carbon taxes, but I’ve never seen an argument that carbon taxes are not going to work because of intermittency.”
Knowledge at Wharton: That’s why we have a worldwide grid because the sun’s always shining someplace, right?
Netessine: That’s right. A grid operator basically borrows electricity from somewhere, but it can be done in a smart way. For example, you have a three-hour difference between the eastern and western United States, so the high consumption period is a little staggered. You could play with that. But you cannot completely get rid of gas-fired turbines because they’re just so flexible, so nice.
Knowledge at Wharton: Did the conclusions surprise you because they were counterintuitive in a way?
Netessine: Yes. We are not aware of any other work at the time, and this paper took some time to write and publish. But I really have not seen any other articles pointing at this, particularly about renewable energy sources as intermittent. There are many arguments against carbon taxes, but I’ve never seen an argument that carbon taxes are not going to work because of intermittency. I was speaking with three U.S. senators’ offices about this issue, and they mentioned that there was something somewhere in some paper, there was some discussion recently of this intermittency problem, but I’ve yet to find any other references. So yeah, to us it was a bit surprising.
Knowledge at Wharton: Can you go into more detail about the practical implications of this research?
Netessine: In the paper, we looked at several ways to stimulate investment into renewable energy sources. One widely known one is feed-in tariffs, which is to say that you tell the generator of renewable energy that you’re going to pay a fixed tariff for electricity for the next 20 years, for example. This is something that is extensively used in Europe. In the paper, we show that this is not a bad approach. But it’s not going to work as well as dealing with intermittency directly. From a practical standpoint, we argued that you have to somehow provide incentive to electricity generators to reduce intermittency, probably by the means of pooling different energy sources from different technologies that are not correlated strongly with each other. That’s a short-term implication; you have to provide those kinds of incentives.
The longer term is thinking about storage technologies. You can make money nowadays using storage because sometimes a supply for electricity greatly exceeds demand, and for brief moments of time there are negative prices for electricity. You can observe them in the market. The operator of the grid is begging you take electricity and pays you take it, right? That’s bizarre. That should not be happening in theory. If you have some kind of industrial-sized storage technology, you can actually make a lot of money during those time periods. You can basically buy energy for almost nothing and then give it back when the price is particularly high, when demand greatly exceeds the supply.
There are a few ideas for storage, but they’re all at the prototyping stage as far as I can tell. I think a part of the reason is that there has never been a recognition that you need storage with renewable energy sources. Imagine the situation where half of your energy needs are renewable, which are intermittent and go in and out, then you have a huge demand for storage. You would think companies would spend more money researching energy storage, and I just don’t think enough money is spent on that. Tesla is developing energy storage in batteries, but everything else is in the prototype stage as far as I know.
“From a practical standpoint, we argued that you have to somehow provide incentive to electricity generators to reduce intermittency.”
Knowledge at Wharton: It has to start somewhere.
Netessine: Yes, and the problem is the prices for batteries for storage keep falling. Once Tesla starts production in its huge factory, it’s going to fall even further. Then once it gets some experience, it’s going to fall even further. You could see how people are hesitant to invest now when the cost of storage is high, which is normal with new technology. Somebody’s going to lead the way, and it’s probably going to be Tesla.
Knowledge at Wharton: What else should we know about this paper?
Netessine: Another aspect that we studied in the paper is deregulation. Deregulation happened in the U.S. market, and there are some benefits of deregulation. But when it comes to renewable energy sources, what we find is that deregulation amplifies this problem with intermittency. Even if everything was perfectly integrated and there was one government running everything, you would still have this issue that you need backup generation to deal with the renewables. But when industry is deregulated, the issue is amplified. There is going to be more under-investment in renewable energy sources under competitions and under vertical integration.
Knowledge at Wharton: Why does that happen?
Netessine: It’s because of the way electricity is priced in the market. When demand greatly exceeds supply, prices for electricity are high. Imagine there is a grid operator who can use hydroelectricity, hydropower, solar, nuclear and so on. What tends to happen first is the grid operator uses renewable energy sources because they have zero marginal cost. That’s the optimal thing to do, but this is when prices are the lowest. So, the first part of demand is covered by renewable energy sources and you don’t get such a high price for your electricity. On the other hand, when demand is huge and these backup generators are brought online, the person who runs this kind of backup generation benefits the most even though he’s going to be taxed the most on the carbon. But the prices are also very high. It turns out that under this kind of a system, which was designed to accommodate all these competing suppliers of electricity, you might over-invest into non-renewable energy sources.
Knowledge at Wharton: Ultimately, we need storage to make it all work.
Netessine: Yes, probably some kind of a combination of energy storage and pooling of various sources of a renewable energy. That’s ultimately what we’re going to need. I think if you give some kind of incentives to generators to be less intermittent, to provide electricity in a more stable way, then they will probably start investing in R&D first for energy storage and ultimately into energy-storage solutions.