Rahul Tongia led the Brookings India sustainability practice which recently has transitioned into the Center for Social and Economic Progress. Dr. Tongia is also a founder of the India smart grid forum and has written extensively on the system level challenges associated with the grid integration of Renewable Energy sources.
In this conversation, Dr. Tongia expands on this topic, and breaks down the system levels costs associated with renewable energy integration into its various components. Hope you enjoy my conversation with Dr. Tongia!
Topics covered in this podcast: 1. How does the level of RE integration in India compare with other countries? 2. Will high RE penetration levels pose significant challenges to grid operators? 3. Is there a penetration threshold for renewable energy onto the grid, which once exceeded, grid operators will then face great difficulty to manage the energy flows? 4. A detailed breakdown of the system level impacts of Renewable energy on the grid 5. Analysis of previously calculated system level costs of RE integration, and Dr. Tongia's perspective on whether these estimated cost quantities are representative of true costs.
00:06 Karan Takhar
Hello everyone, this is Karan Takhar, and welcome to the energy podcast. Over the past decade, India has done an impressive job of integrating renewable energy into its energy mix. For this Fulbright Podcast series, I sought to investigate the enabling factors and potential of India's global leadership in renewable energy, with the focus on solar. This Fulbright series is broken down into Four Seasons. This season, we look at the next set of key technologies and regulations integral for unlocking India's continued renewable energy success at the system level. It includes conversations with leading regulators and thought leaders across energy management, storage, transmission, and distribution. In this episode, I will be speaking with Rahul Tongia, who leads the energy and Sustainability group at Brookings India. Doctor Tongia helps found the smart grid task force of the Government of India as well as the India Smart Grid Forum in multistakeholder body chaired by the Ministry of Power. Doctor Tongia is a leading voice in quantifying the system-level costs of renewable energy integration into the grid, and in this conversation, we discuss some of his views on this subject. I hope you enjoy my conversation with Doctor Tongia.
01:50 Rahul Tongia
I think this is an interesting and complex space, and while I've been trying to learn for several years on some of the nuances, I think the bottom line is nobody exactly knows because the first question that comes to mind is this any different for India than any other country which is aiming to expand its renewable energy portfolio, its basket, and renewables has three have three challenges worldwide. The first is, in general, people used to believe they're expensive. That's not necessarily the case anymore. The second thing is they're variable. So this is called variable Ari. Take it when you can when the sun is shining or the wind is blowing and the third one is there location dependence. Solar is relatively better spread, but wind especially is very location specific in India, the US, and most other countries India has not yet. For example, that much offshore wind and so the first question is what is different or special about India and the quick sort of high-level thoughts are why, well, every country is different but India has a couple of characteristics that distinguish it or separate it from many of its peers or similar countries along their transition trajectory. 1st is the Indian grid was traditionally very coal-heavy, somewhere in the order of 3/4 last year of electricity generated was from coal. The second was Indian grid has been relatively weak compared to many other grids. Now, there are many metrics and measures of this, but until recently there were shortfalls in supply. You actually had a lot more load shedding, as it's called, which is cutting off demand just because you don't have enough supply. That problem has diminished. It has not gone away, but it was also a grid that was reasonably light in terms of its ability to have buffers and surplus and things like this. So even though now India has had for several years a National Grid just like the US. A National Grid does not mean one can effortlessly send power from one end to the other infinitely. There are interconnections between regions, but they're not very strong, the tie lines as they're called, and that's the same in the US where you have 3 synchronous areas, Eastern, Western, and Irkut, or Texas. But that doesn't mean you can send any amount of power from Oklahoma to Maine or vice versa. There are limits to that and those limits or even sharper in India, where a lot of dispatch decisions are taken at a state level, but the interconnections are strong enough up to a regional level. So these are some of the high-level thoughts on renewables. Another small one is that a lot of the growth. Is in solar power, not wind. That's probably true in much of the world, but unlike say Germany or California, well, or at least parts of the USA lot or if not a most overwhelming amount of the growth in solar power has come from grid-scale solar parts or solar farms. These. Or not rooftop feeding into the local grid. It happens in India, but much less. Then in some countries, proportionately, and also much less than the targets, have asked for so-called rooftop or behind-the-meter renewables. That's another characteristic which has its own implications, both technical and financial.
05:37 Karan Takhar
I see, And I was watching a talk that you delivered at the India Smart Utility week, titled Mainstreaming Renewables, and this might be difficult to put a precise like yes or no onto, but so you stated that having a small fraction of the grid provided by renewable energy is relatively easy. Now what that number is will vary across foods, across countries, across domains. But my question is do you believe like this threshold is being reached in India where like that small fraction which was easy to control is now being exceeded and we're at the stage where.
06:20 Rahul Tongia
06:22 Karan Takhar
Like the threshold limit has been passed percent, where now renewables are being integrated more heavily and it's causing a lot more trouble at the grid level.
06:34 Rahul Tongia
So the good news is I believe we're not anywhere close to problematic areas.
06:40 Karan Takhar
06:41 Rahul Tongia
Quantum of renewables thus far is modest, and even if we look at the target of 175 gig watts, which may or may not happen by 2022. I don't want to comment on that specifically, but in my mind, it's OK if it doesn't happen by 2022. It may just happen a little bit later. It's still very ambitious and admirable India targets just to give context on those targets. You have countries, you have states saying I'll become low carbon or carbon 0 in electricity, or maybe even in all energy. That's one important wants by 20, thirty 2040, or 2050. What that really means is for a developed region, you don't have too much growth of overall energy, you've already got very high consumptions and so all you need to achieve those Is maybe, let's say if you have 30 years to meet that target, you just need to switch over 3.33% every year, assuming you're at zero, and in 30 years you'll hit 100% renewables. But for a place like India, the targets announced in late 2014, early 2015. I mean, we're so high, the 175 gig watts that that translated to a 25% annual growth of renewable energy, which is astonishingly high. So even if those targets aren't met in 2022, that's OK. Those numbers are themselves probably manageable. It's a little bit after that when you start to need more fundamental changes in how you look at things. The first would be grid operations that are signaling so that you recognize different electricity's are not the same time of day. Pricing is just one of it And the second is the eventual need for storage technologies. So, coming to your first point. How much can you handle it's a spectrum. It's a continuum. It's not as if it's easy or hard. It's just easier or harder, and one of the truisms of renewable energy is as you grow the share of VRE in your system, any system worldwide. Two things happen, and it's a gradual process. The first is the marginal utility diminishes, so think of it this way. As I add renewables, which are relatively cheap, I'm going to knock off more expensive alternatives. But once the low-hanging fruit is done, and the lowest of easiest to knock off is the variable cost of running existing coal plants and more for the fuel costs of coal, if I switch off my coal plant, I don't save anything on building it 'cause. It's already built, but now I save on fuel. That's a very easy one is to one comparison but once I stop being able to displace expensive cold or mid-price coal, there is still a lot of cheap coal. Which is often run off mine or within India, is called pithead cold, where you don't have transportation costs. That is much cheaper and harder for renewables to displays at the margin. The second related point on renewables is as you grow more and more, the marginal cost of integrating them rises. So you will eventually need more transmission because whatever you can absorb nearby is done because of this time of day dependence of renewables, it peaks solar in the middle of the day, for example, that's very intuitive and obvious. So you first try and absorb it locally, state-level regionally, but once you have quote unquote too much, you have to ship it elsewhere and at some point either you have to add a lot more transmission and after that point, even if I had infinite transmission at a national level, my demand and supply may have a mismatch.
10:30 Karan Takhar
OK, that makes sense. Is there, is it possible to like provide a potential range in terms of like this percentage of the grid? If renewables accounted for this by the percentage of the energy mix, then that's when there's that It's just you can't really put a number on it.
10:53 Rahul Tongia
It's hard to put a number in part because there are several moving parts. If my grid looked identical to what it does today, meaning the rest of my grid is well known. The rest of technologies like storage and demand response smart grid are known and might load profiles.
11:06 Karan Takhar
11:13 Rahul Tongia
Which include the shape at which I have demand are static. Then it's easier to make that claim. But there's a feedback loop because just like time of day could impact the value of different generators. So if I have a whole mess of solar. A whole bunch of solar, then that diminishes the value of that solar because all of it becomes available at around noon. But my demand can also respond the time of day signaling. But today's demand peaks in the evening. It's very residential driven, air conditioning load driven, and so it's often between 7:30 and 11:00 PM when solar output is near 0 but what's to say that at least some of our demand can't be shifted through control mechanisms, pricing mechanisms, policy incentives, a behavioral shift, etc, and so it's very hard to come up with that number though the 175 GW target translated to assuming a certain growth of demand between 2017 and 2020, which is when we've done some calculations. It was in the order of 20% coming from Ari 20% seems easy. You can manage the 20% quite easily because you have so-called surplus coal, but in the middle of the day, all you do is you switch off or you lower the outputs. You don't actually switch off a coal plant. It's very expensive to switch on and off a coal plant, but what you do is you lower the output. Any given a certain coal plant, what's called coal flexible operations or flexibilization or flexing and you come down to not 100%, but typically the target is 55% as per government norm. So one of the masks that you can do is ask yourself, how much does a 55% translate, and because you still need the coal plants to turn on for the night. So unless you had storage and your demand profile doesn't shift, then you could get an estimate that OK in the evening I gotta have this much supply room solar is not going to cut it, so I need to have my coal plants back at 100% technologically you can go from low to high over a span of a few hours, that's not hard technologically, and therefore it's quote unquote. Doable, but it also has a small cost because when you ramp your coal plants up and down, it has wear and tear, and operational impacts, and it also lowers their efficiency when they don't operate at 100% loading. But that's OK. That's still a minor hit. It's quantifiable, it's justifiable, it's manageable, but now crude estimates indicate that you only have maybe in the order of another 30,40,45 gigawatts of flex capability. So that's the additional VRE that I could have mid-day hypothetic. Before Rich, I need major changes or new investments in alternative generation. So the 1st next rung on the ladder would be when you need to start to have quote-unquote something else the easiest is when I don't need to build anything new other than Ari, and I can just manage in my existing grid. Is the rest of it is going to be Insurmountable, No. It's going to be gradual, increasing over time. How much battery would I need? Well, I only need a little bit, because it's only going to happen some days of the year, some hours of the day, etc. But over time, as your renewables grow further, you're going to need more and more.
14:47 Karan Takhar
So as I was reading the report each put out on the complexities of integrating Ari into India grid, or you write that the LCOE. Unlike comparing fuels or energy sources based on, LCOE isn't fully comprehensive because renewable energy has an added system-level cost ending one could you talk a little bit more in detail about the system-level costs of Ari as well as so are you recommending that these are costs like which should be borne by the generator or was the intent mostly to just bring awareness to these impacts, as I was reading it, I was curious about that.
15:32 Rahul Tongia
It's not just awareness. I'll come back to this after your first question, but we need.
15:36 Karan Takhar
15:36 Rahul Tongia
A way of quantifying it. So the short answer to your second question is we have to quantify what those costs are and they're not well quantified because the current policies are ones of socialization. Whatever it is, it's not signaled we'll just put it into total costs paid by everyone on average. So one classic example of that is transmission let's say hypothetically I need dedicated transmission lines, and the most extreme example of that is point-to-point. So in a point-to-point system, I know exactly what it costs if we make an assumption that this dedicated line is only used or loaded partly because, for example, solar utilization is only 25% capacity utilization factor or in India called fleet load factor. So it's less than many other transmission lines, and therefore it will be more expensive than your average, but you don't charge it higher. What happens today for most transmission is the total transmission for an area, whether it's state level or intercity. For regional their different mechanisms is total transmission divided by total kilowatt hours and you just charge. That amount there are forced locational nuances and differences, but it's still a high level of socialization or averaging going on, so that's the first thing that we have to think about more precise costing of with system level cost, you mentioned flexing of coal plants. It's not a stranded asset necessarily. I mean, the extreme example of stranded is where you're paying for capacity that doesn't get used here It would be more like a spectrum because you're underutilizing it to its potential. So if I build a coal power plant and let's assume its efficiency is static. It doesn't change based on loading, and it has a certain fixed price of fuel. We know what its goal costs and now I use the same coal power plant for 50% of the time or 80% of the time I'm assuming no coal plant actually runs at 100% loading over the year. It has to have making experience. Downtime, so let's say I have a plant that operates at 40 and are not another plant that operates at 80 round numbers. So my variable cost in this hypothetical example is identical. My fixed costs are amortized on a base that's twice as large in one case as the other, so my average cost looks different. Now that's the point of looking at costs on a portfolio basis because if I only looked at the marginal cost, it doesn't change and now you would say, hey, it's cheaper to run my Ari, so that's a second calculation if I either already have or have to have something else, quote unquote other than the renewables, and now I say my renewables can generate, let's pick a number rupees 2.36, which is the new low it was discovered very recently 2.36 is pretty cheap, and it's actually cheaper than the fuel cost at a bunch of coal plants. Not all very location specific, but a bunch of them and so 2.36. Reduces your variable cost out of the coal, which gets back down, which gets told to flex down or back down it couldn't produce. What you're asking it not to be at a grid dispatch level. But I'm still paying the fixed costs of those coal plants that are needed because they either they were already built and have to be paid off and or even if it's there someone paid for it, even if they paid for it in the past. So the way that you handle this is you need better signaling for what is the value of increment. So supply into your mix at any given point in time. So that's starting to sound like an electricity market. It doesn't have to be a market mechanism per se. There are many types of markets. Most people think of the energy market, and electricity markets as being energy markets. You could have capacity markets as well, but more than that you could even just say. I will take current bilateral contracts of energy and modify them to add a time-of-day adjustment for what it's worth to have so most power purchase agreements. The bilaterals are costs. In cost plus, all we do is we make an assumption of how much it's going to operate. So there's a normative plant load factor usually, 68.5%, which is where they do their economics that assuming it runs this much, this is what you have to pay and depending on the type of contract if the state uses the power plant. By their choice, maybe demand didn't grow enough or there's something else. They are still on the hook and obligated to pay the fixed costs, assuming that the power plant was available to generate but it backed down due to no fault of the generator. So in that case there's no risk, but the state is still on the hook for paying those fixed costs. So what we really need is a way to signal how much of what I need when. So if I really had a true nimble market and at one extreme it's an energy-only market. One theory says that people should bid at their marginal cost. Which means as long as I get paid more than my fuel, I'll be happy to run. Now obviously, if I only got paid my marginal cost, how would I recover my fixed cost. So the assumption is in a uniform price market, the last market cleared price is higher than the marginal cost of many of the cheaper units, but they get extra revenue, and that covers. Their fixed costs that’s the general norm and it has some variances. But it often plays out, but there are limitations to that one model, but now think about places which have a lot of supply and very little demand prices fall and in fact, in places in the US, Germany, you have seen negative prices for electricity. Now just think about that. Doesn't that blow your mind, What it really is telling you, we've got temporary periods where there's just way too much supply, and please use it. I'll pay you to use electricity.
22:06 Karan Takhar
Wow. Yeah, that's interesting.
22:08 Rahul Tongia
Doesn't happen too much, but it signals one of the limitations that's going to become more and more stark because the marginal cost of renewables is 0 once you've built them the incremental cost of running them is usually near 0.
22:29 Karan Takhar
I now see a recently released delivered report which estimates the system level impact of Ari to be about rupees 1.11 per kWh across four different charge categories, including the total balancing charge, the impact on TSM per unit standby charge, and extra transmission charge and then so that was rupees 1.11 kilowatt per hour, and then as the chair, if you Kirk, he referenced the CA analysis of the rupees 1.11 but then added an additional charge and he calculated the charge that results from that drop in PLS to be rupees 1.05 per kWh.
23:20 Rahul Tongia
So a couple of things.
23:14 Karan Takhar
Yeah, I'd love to hear your thoughts on this essentially.
23:17 Rahul Tongia
I think it's we can now go back and step look at the other side of the coin.
23:24 Karan Takhar
23:25 Rahul Tongia
The whole grid should have been a lot more nimble with a lot less locking and a lot less overcapacity, so. It's unfair to blame renewables entirely for this. The falling PLS. What happened was between 2011 and 2016, the growth in capacity of coal power plants in India was somewhere close to 14.5% annually, and in the same period, the demand for electricity only grew at about I think 6.6 percent or something like that, and so there was a huge growth in coal power plant capacity. Now that is the obvious reason why you have a fall in PLS. Sure, renewables grew over this period. I'm going to guess from just a couple of percent to 8%, somewhere in that range 7%. So yes, it did grow, but a lot more of that fall Is because of just the sheer growth of capacity. So you have done a very poor job of signaling what capacity. So if I step back and I say the peak is your problem in India, and the peak is in the evening if those two facts remain true. Then you don't want to add more solar because it doesn't help you for the peak. Not without a battery, but at the same time, you don't want to go out of a new coal power plant as well and then say, oops, I or I needed this cold path right now. Please pay me because I got to recover all these fixed costs. That's also bad design.
25:03 Karan Takhar
That makes sense.
25:05 Rahul Tongia
So the larger failing was also that coal power plants just assumed that this static world with fixed costs and high PLF that are predictable will continue forever, and when you're in a cost-plus regime, the only thing you worry about is not efficiency per se, but getting an approval for your rate base or your power purchase agreement and that's not very good for the consumers either. So we do want coal power plants designed such that they can flex more and more and in fact, a lot of the plants were asked to be so but didn't really comply well or fail to do it properly, and again, I'm not blaming any single plant or type of plant newer plants are better than older, but we should in general have been doing a much better job of it across the spectrum, so I also would counter that just because someone says let's make up a number 1.11 rupees a kWh. It doesn't mean. That all of the those costs are acceptable or should be passed on to consumers. Some of them should be borne by different entities as well because if we go back to the traditional form of regulation, once the cost is approved, that it's not my fault, then it has to be borne by the consumers, my general OK, so I'm going to cheat, I'm going to give the answer, I told all my students the answer to any policy question is it depends your job is to now figure out what does. It depend on how and why, and so to try and come up with one number is also misleading, because that number will change by season, by time of day, what the rest of your grid looks like, how you price, other factors, etc. At best you can easily come up with an average number that reflects today, and both of those can and should change, and the good news is we're getting smarter about how we make our grid more nimble, our reflects down more without risking stability. How we add more maybe storage or demand response, these are also getting cheaper. So the good news is I don't think that' a number is going to just be as bad as people think. The flip side is as renewables grow. It's going to displace obviously the more expensive coal first, but at some point you're going to then get rid of all of your expensive coal plants, at least at the margin, meaning they won't operate as much. It's unlikely that you'll dynamite them. You still may use them, but less so, but at some point. You'll run out of these lower-hanging fruits and you'll have to compete with variable costs of one of my, one of mine, or pithead power. But at some point, even assuming it's quite cheap, even assuming it beats every single coal plant in the country hypothetically, there's infinite transmission, etc. Even then, at some point, you will come to this next rung on the ladder, which says, now what do. Hey, honey. I do at 8:00 PM so you're going to have to start to include more and more storage capacity as well. My largest takeaway is this stuff is a process. We shouldn't get bogged down with any specific number because I can guarantee you two years later that number is going to be wrong. OK, But we should be more cognizant of the trends across over points and what are any things that are nonlinear or discontinuous. So for example, today the variable costs to manage our Ebooks are low, in part because I have surplus coal. But if I ever run out of surplus coal, then I'm going to have to make a call on what else do I invest in. Do I want to build a new coal plant? Do I want to maybe build a gas plant do I want to the hydro? Do I want to build quote unquote an expensive battery because obviously if their batteries were dirt cheap, we wouldn't be having this conversation? A lot of so in a couple of weeks, we'll be releasing our book on the future of coal in India and you'll find a lot more nuance and depth in that book.
29:20 Karan Takhar
Oh, amazing. Well, it'll be released on the Brookings website.
29:26 Rahul Tongia
Well, the link to it I mean the publisher will have copies and we'll have excerpts on our website.
29:31 Karan Takhar
Done. Yeah, I'll go. I'm definitely gonna get a copy. Thank you, doctor Tongia. Sincerely and.
29:37 Rahul Tongia
Yeah, cheers. Current my pleasure.
29:40 Karan Takhar
You know, I'll keep you posted. Thank you very much.
29:40 Rahul Tongia
29:44 Karan Takhar
I hope you enjoyed that episode and do check out the show notes For more information on my guest. See you next time.