The Indian Power Grid: If Renewables are the Answer, what was the Question? RAHUL TONGIA1 ndia has ambitious goals for renewable power, and India needs to manage a while solar power has garnered much of the excite- ment (and recent investments), all forms of green capacity problem, unlike Ipower have Central government support. End-users? many countries seeing RE They worry about their lights being on (i.e., load-shed- ding), and while they might be philosophically support- as an energy solution ive of green power, it’s unclear how much premium they How do other countries handle large RE inputs to the may be willing to pay for renewable energy (RE). This grid? They do what the power industry has always done leaves the electricity utilities as key stakeholders, with – engineer for contingencies, with reserves and alterna- different roles spanning generation, retail, distribution, tives. Unfortunately, India is short of power, and thus, transmission, and balancing the grid. Most states in while other nations might see RE as helping an energy India have unbundled the power sector (separating gen- (kWh) problem, India also needs to manage a capacity eration, transmission, and distribution), but regardless (kW) problem, which is precisely where RE falls short. of institutional structure and reforms (or even state of Renewables in India are different from renewables de- finances, a critical topic), all power grids require bal- ployed in the U.S., Europe, etc. and understanding ancing, i.e., continuously keeping supply in sync with these differences is key to viable policies. The triad of varying demand (net of losses) in real time. This is the “usual” challenges of renewables remains in India, such basic principle of the modern alternating current (AC) as (1) intermittency/variability; (2) location-specific grid, which has a characteristic that electricity cannot potential (concentrated in areas sometimes away from be stored easily. consumers or the grid; and (3) higher costs. Howev- er, there are specific differences and needs that demand This is where the fundamental variability and stochas- deeper analysis for the long-term viability of renew- ticity (randomness) of renewables become a challenge, able energy. Making renewables viable for producers is more so for the Indian grid which is operated far from easy—pay them enough—but can the rest of the system optimally (with outages and load-shedding, frequency handle that? Because of pricing subsidies as well as high drifts, etc.) losses (both technical and commercial, i.e., theft), util- 1 Portions of this chapter draw or are taken from a PlanetPolicy blog piece done by the author (http://www.brookings.edu/blogs/planetpolicy/posts/2014/05/21-renew- able-energy-india-tongia). 22 | Blowing Hard or Shining Bright? Making Renewable Power Sustainable in India ities already lose on average about a rupee, if not more, More than just operations, even planning for RE inte- per kilowatt hour sold, impacting the ability to finance gration is important, including coordination between all new generation capacity, just RE. the various stakeholders. At a 2011 Workshop on Wind Power in Karnataka, Karnataka Renewable En- One of the typical calculations that power system op- ergy Development Ltd. (KREDL) proudly announced erators do is estimate how much renewable power the they had sanctioned windpower projects in Karnataka grid can handle. Typical figures from elsewhere are in for roughly 12,000 MW of capacity (comparable to the the range of 20-30%,2 with more requiring signifi- total official capacity of the state), and that over a quar- cant investments in transmission or peaker plants. Of ter had achieved financial sanction as well. The one course, with more deployments and learning (and bet- catch, which came to light in the frantic discussions ter predictions), the ability to handle RE is increasing, post KREDL’s presentation, was that nobody from Kar- and, most importantly, every region is different. India’s nataka Power Transmission Corporation Ltd. (KPTCL) grid is weak and unstable, and instead of having a rea- (which was engaging with KREDL for the first time sonable reserve margin (typically 15-20% in the west), on such issues at the Workshop) was aware of any such there is a shortfall in the grid, officially in the range developments, not even from a power evacuation per- of 5% or so, but actually much higher. Even the Grid spective (availability of substations), let alone issues re- Code is modest, recommending (but not mandating) lating to handling the variability of windpower. The only a 5% margin. The grid is kept afloat through mas- challenge, fundamentally, was one where the “makers” sive “load-shedding” (feeder-level supply cuts). and “takers” of renewables didn’t coordinate. There are other technical reasons why the Indian grid is Overall improvements in the weak, including lack of ancillary services (systems de- signed to keep the grid stable, instead of just pricing grid, including better kilowatt-hours), and even a lack of time-of-day pricing balancing without resorting for bulk procurement of power. There are few peak- er plants (which would operate only some 5-10% of to load-shedding, should hours in a year), since there isn’t sufficient incentive for be key areas of effort, which these. Without incentives for plants that can ramp up (or down) quickly but may not get used much, how would facilitate increased will the grid handle 20% renewables? Even worse, the RE penetration types of plants capable of fast ramping are limited in near-term growth in India—hydropower (due to land GRID DETAILS, FUNDAMENTALS and social/environmental challenges) and natural gas AND NUANCES (due to supply constraints). Hydropower has an addi- tional constraint when considering peaking or storage It’s well known that the Indian peak demand (today – its additional duty for water management (irrigation) and in the near future) is mostly in the evening. While limits when water can be stored versus released. Overall it’s obvious that the sun isn’t shining bright then, even improvements in the grid, including better balancing wind is often on a ramp-down, especially in some states. without resorting to load-shedding, should be key areas Thus, even if India adds 20 GW of solar, it still needs of effort, which would facilitate increased RE penetra- 20 GW of additional capacity to meet its peak, and the tion. picture is almost as bad for wind because of its strong 2 http://energy.gov/eere/wind/renewable-systems-integration. 3 The official shortfall of power in India is now only 5%, but the actual shortfall is far higher. See Brookings India Discussion Paper 01-2014 on Re-thinking Access and Electrification in India: From Wire to Service (http://brookings.in/wp-content/uploads/2014/09/electrification-from-wire-to-service.pdf). Blowing Hard or Shining Bright? Making Renewable Power Sustainable in India | 23 seasonality. Actually, it’s not quite as bleak, since RE While the annual average Plant Load Factor (PLF) is offers some capacity value, but the exact amount it of- just over 22%, not only is there strong seasonal varia- fers is modest vis-à-vis conventional fuels, and we need tion, even during windy periods the output falls to near detailed calculations for these at a systems level, instead zero. Also, at least for Karnataka, the fact there are sev- of simplified correlation studies. Only for illustrative eral windy zones separated by hundreds of kilometers purposes, the chapter on Economics of Renewables doesn’t steady the aggregate wind output very much. compares different pricing regimes factoring in the sep- Converting these into month-wise PLFs, we clearly see aration of energy and capacity. the enormous variation. If the Indian grid had buffer ca- pacity, and there were better regional interconnections As we can see for the example of Karnataka (which with non-coincident peaks, then we could plan more may or may not match other states), there is enormous in terms of RE and demand coincidence, like Germany variance, on a nameplate capacity that started 2012 at and other countries, but as of now, even being opportu- 1,851 MW rising to 2,168 MW by the end of 2013. nistic with RE, we still need more peak capacity. FIGURE 1: Total Wind Power Output in Karnataka at 1-Minute Resolution 2,500 2,000 2012 Wind Capacity 1,500 2013 Wind Capacity 2013 Wind Output 2012 Wind Output MW 1,000 500 0 Jan Feb Mar April May Jun July Aug Sep Oct Nov Dec SOURCE: KPTCL Raw Data 24 | Blowing Hard or Shining Bright? Making Renewable Power Sustainable in India FIGURE 2: Month-wise Total Windpower Plant Load Factor (Capacity Factor) in Karnataka. These are calculated from 1 minute data across the total state. The annual average PLF in both years is just over 22% (22.19% and 22.46%, respectively). 60% 50% 2012 2013 40% 30% 20% 10% 0% 1 2 3 4 5 6 7 8 9 10 11 12 Month SOURCE: KPTCL Data Variance is one reason why renewables aren’t a panacea Balancing the Grid for rural electrification. Beyond the issue of the evening peak (demanding a battery if standalone RE or an iso- Much more subtle than the challenge of meeting the lated microgrid is to provide lighting), most optimal peak is keeping the grid in balance. Germany now han- renewables (except solar) are village-scale, if not larger, dles 25% RE without major investments into the grid, not household. One still needs a last-mile connection. but the German Energy Agency (DENA) estimates At that point, the grid becomes more attractive (given it further increases will need large investments in grid SOURCE:SOURCE: KPTCL KPTCL Data Data reaches the vast majority of villages already), especially strengthening, especially high voltage transmission (re- as demand grows rapidly once a household is electrified.