Can a Coal Power Plant Ever be Good?

April 22, 2008 | Siddhartha Shome,

Recently, plans for a new "Ultra Mega" 4,000MW coal fired power plant in India has come in for much criticism from environmentalists. Writing on, environmentalist Nathan Wyeth has called this a "monument to a failed approach". According to him,

Investing in coal generation and plugging it into an unreliable grid (rather than building renewables close to consumers or fixing the grid) has the effect of - get ready for this - spurring the construction of small-scale fossil fuel generation on the other end, which is ... incredibly dirty.

In my opinion, Wyeth's analysis is flawed. It does not take into consideration how, and for what purpose, small-scale fossil fuel generation is widely used in India. Let me give you a real-life example.

The company I work for has a software development center in Pune, India, which I occasionally visit. It is a large modern office building, in many ways better than our office here in Silicon Valley, California. However, there is one problem: power cuts. In order to deal with these, our company has installed back-up diesel generators - generators powerful enough to operate the air-conditioning system, the elevators, the lights, the computers, etc. While I was there, power cuts would occur approximately every other day for a couple of hours. The diesel generators would immediately swing into action, providing uninterrupted electric supply, but also spewing out huge quantities of thick black smoke into the city air.

This is surely the worst possible way to generate electricity. It is much worse than a large modern coal power plant, not only in terms of CO2 but even in terms of sheer "pollution" - the odor, the particulate emission, etc. As in my company, the most common reason for installing small-scale diesel generators in India is the need to have a back-up power supply. These diesel generators are expensive to operate, and are used only when absolutely necessary. How long they operate (and how much dirty emissions they produce) therefore corresponds directly to the deficit between the electricity demand and the (grid) supply. If a new power plant - even a coal fired one - can reduce the usage of these small-scale power generators, it will likely reduce carbon emissions overall for a given electricity demand.

Of course, the problems with the electricity grid in India are part of a larger problem of underdeveloped infrastructure - something we take for granted here in the U.S. Governance and administration at local levels have failed to provide adequate utilities and public services, such as electricity distribution, water, maintenance of roads and public transport, etc. There is certainly a very real need for investments in improving the electricity grid in India. However, it would be foolish to take the position that India should not build any new electricity generation capacity until all the flaws in the electricity grid are completely ironed out.

Whatever problems there are in the electricity grid, these problems are equally applicable to some forms renewable energy as well. Wind power is making headway in India, and one of the world's largest wind power companies, Suzlon Energy, is an Indian one. But wind power also needs the grid to deliver electricity to consumers since wind farms tend to be located far from population centers.

The grid losses that Wyeth cites are not as bad as they appear at first glance. True, the statistics on grid losses in India are staggering. However, it is not that all this electricity is somehow vanishing into thin air. The bulk of these losses come from people actually utilizing the electricity and not paying for it. Sometimes, this is done by unscrupulous individuals. However, in many cases, this "lost" electricity is put to perfectly good use. For example, many people living in shanty-towns in India do not have formal electricity connections. Local politicians eager to keep voters happy sometimes arrange for informal electricity connections. This electricity is counted as a "loss" in all the statistics, but is actually put to good use. No doubt, this is far from an ideal arrangement, and cannot substitute for proper commercial arrangements - but it goes to show that not all grid losses should be seen as equally wasteful.

Moving forward, it is clear that the Indian electricity grid needs to be dramatically improved. What is also clear that with a growing population and a growing economy, the legitimate demand for electricity is going to grow by leaps and bounds, and all that new electricity generation capacity has to come from somewhere. The question is: what technologies should be used? I would rank energy generation technologies in order of preference as follows:

1.Renewables (wind, solar, etc.)
2.Clean non-renewables (nuclear)
3.Moderately dirty but efficient fossil fuel power plants (e.g., coal-fired plants like the Tata Ultra Mega plant)
4.Small scale inefficient and extremely dirty generators powered by diesel or other fossil fuels.

Ideally I would like all electricity to be generated by renewables (who wouldn't?). But as we all know, this is not commercially viable today. Insofar as a coal-fired power plant replaces forms of power generation that are far dirtier, like diesel generators, and make electricity available to people without electricity, a relatively efficient coal-fired power plant, such as the Tata Ultra Mega plant, should be seen as a good thing.


Thanks for this good post....

By Solar Products on 2009 06 24

Vikash mentioned negawatts above, something that seems to have been left out of the initial discussion about RE v Coal...Here are some suggestions which were recently sent around Mumbai.

*Switch AC on an hour after coming to work and an hour before leaving.
*Run all ACs at 24 C (3%-5% energy saving for each degree above the common 22 C practice.
*Switch off from the plug switch (5% of usage)
*Use the sleep and hibernation mode on computers (turning them off all together when not in use is an even better option)
*Replace incandescent with compact fluorescent lamps (LED saves even more energy)

So the energy providers are not unaware of this problem of power outages and they even understand the economics. The economics are that the provider loses potential revenue from these outages. The fee structure of energy tariffs in India makes these losses worse as companies pay almost double that of residential and agricultural users. The power company has begun to realize that by promoting negawatts they can increase revenue without expensive investments in infrastructure.

Now the question is why is TATA producing a massive new coal power plant instead of distributed renewable systems...Mostly because coal is the most cost-effective way to solve the energy crisis now. But will this new generation really help remove the need for the diesel generators (remember that the need for Electricity is projected to grow at ~10,000 MW per year)?

Not really, since taking into account the 25-40% T&D losses makes this really a 2400-3000 MW power plant. The infrastructure for electricity delivery is so out-dated that the entire system would have to be revamped and almost replaced, just to make the addition of generation available to the end-user. The fact remains that the last mile is non-judgmental. All energy sources are exposed to the same losses in the last mile...but the last mile doesn't have to exist. There are options to create localized privately run distributed systems which are not grid connected and thus not exposed to the last mile. The size of these units would be much smaller than those proposed above, but done in the proper way...maybe the grid could become the source of last resort.

Okay pipe dream, this option is not fully financially viable, yet with the rising costs of diesel long will diesel be a viable alternative?

My point is that the power outages are not going to go away in the near future...The price of diesel is going to rise...a diesel based system is no longer the 'best, cheapest, most cost-effective available technology'...time for business to pull its head out of the sand.

By Timothy Fox on 2008 05 23

Close to 90% of existing power consumption is wasteful. Demand-side Management or Negawatts as they are called, holds the key to our survival and prosperity. Here are some pointers :

1. Full utilisation of daylight ( provided by a super-ultra-mega fusion power plant 93000 miles away) Why do we need all the night lights, and all the light pollution anyway?

2. Good ventilation and airflow design in neighborhoods/blocks/buildings combined with greenery and waterbodies; wholehouse fans, heat sinks combined with solar-thermal, micro-wind AND internal power and conditioning such as Yoga or other meditation which increases our comfort range of temperature can eliminate the need for room air-conditioning, water heating/cooling even with 5 star energy saver ratings.

3. Relocalization, good organisational and workflow design can curtail the need to keep many of the machines to be power-on upto 50%.

4. Renewables if properly invested in, can cheap in to meet the 10% demand.

5. Let us leave coal,hydrates,uranium ( and all the fossil-fuel needed to produce fissile material,build and run nuclear-power plants) in the ground or below the ocean, so that we do not destroy the critical resource of relative climate stability - key to our health and nutrition.

Overlooking the rigors of inculcating a new culture of power-lite, high internal energy lifestyle, will take us through endless debates to virtually nowhere.


By Chandra Vikash on 2008 05 01

Hi Nathan,

Thank you very much for this very well thought out and detailed response. Here are some responses.

1. Your basic assumption is that if 100+MW renewable powerplants can be built "right next to 100+MW demand", these will become cost competitive because the electricity will be transported for shorter distances, and hence grid losses will be less. This is a false assumption. In grid losses - even if you consider technical losses only - almost all the losses happen in the "last mile" stage. See this ( paper that you cited in your Grist post. Look at the table on page 2. This gives a breakup of the various types of technical losses. Only the first item in the table "step-up transformers & EHV transmission system" correspond to losses incurred in long-distance transportation of electricity via high-tension electricity pylons. And guess what? This is a maximum of only 1%. The other categories of technical losses happen in the final step-down transformers and distribution stages, i.e., in the "last-mile" stage. This is just the technical losses. Commercial losses, which anyway form the bulk of the grid losses in India, obviously happen in and around population centers, i.e., in the final distribution or "last-mile" stage, not in the long-distance high-tension electricity transportation stage. In other words, for the Mundra power plant, grid losses will be negligible all the way from the plant to the outskirts of population centers. It will be in the "last-mile" from the outskirts of population centers to the actual consumers where all the grid losses will occur. So building a 100+MW power plant (whether renewable or coal) just outside a population center will, by itself, do nothing to reduce grid losses. Whatever grid losses are expected to occur with the Mundra power plant, the same losses need to be factored in for the 100+MW renewable power plants that you suggest. The only type of renewable technology that can eliminate these "last-mile" grid losses is photovoltaic cells on the roofs of buildings, etc., because only then does the "last-mile" of the grid become redundant. But obviously these are not commercially viable, and I am skeptical of the technical feasibility as well, just considering the limited availability of roof space in office buildings, etc. However way you look at it, the grid looks to be part of the electricity equation for a long time to come, and like so many other technologies, the last mile is where most of the grid problems are. Also consider this simple question: if grid losses could be eliminated simply by building power plants close population centers, why would Tata Power, the World Bank, etc,. not just build 10 400MW coal power plants near population centers rather than a single 4,000MW centralized coal power plant. Building a 400MW is certainly not an insurmountable technical challenge, and the people at Tata Power, the World Bank, etc., are not stupid (at least as far as their own companies' financial interests are concerned).

2. I agree with you 100% that that improving the grid is absolutely necessary, and both technical and commercial losses need to be stemmed. I also agree with you that unaccounted electricity used by poor families is not the biggest problem here. In my opinion, addressing the technical losses is a fairly straightforward issue - its a matter of investing in available and proven technologies. The bigger problem is the issue of commercial losses. Unfortunately, there is a general failure of governance and administration at local levels in India. There are serious problems not only with the electricity grid, but also such basic public services as delivery of household water, maintenance of roads, provision of public health services, public transport, etc. These stem from various political, social, administrative and electoral problems. In India, farmers' electoral clout is very strong (simply because such a large proportion of the population is dependent on agriculture) and so electricity for agriculture is cheap or (in some places) free, even though it is obviously counterproductive in the long term as you point out. I agree entirely with you that it these problems should be addressed. I also agree with you that all these problems make the capital costs of deliverable reliable electricity very high. Where I disagree with you is that you think that these various problems make the capital costs for power high for coal power plants only, but you appear to believe that these problems will not apply to renewable power plants. Renewable generation of electricity will not make all these problems magically disappear. There problems are pretty much independent of the mode of power generation. How do we get farmers in India to pay commercially viable rates for the electricity they use? I really don't know. Maybe political scientists and public policy gurus have some ideas. In this scenario, the logic for large-scale public subsidy for renewable power generation (as opposed to subsidy for renewable technology development, which I think makes sense) becomes even weaker than if these problems hadn't existed. Of course, if the cost of power generation through renewables becomes cheaper than coal power generation, it will be a completely different story.

3. On captive power plants, you yourself say, except for a very few that supply power to massive industrial units (such as the Jamnagar Refinery in Gujarat), the cost of electricity from these is higher than grid electricity. So doesn't it follow logically that users will be tempted to turn these off whenever grid electricity becomes more reliably available. Yes, a 2 hour power outage rather than a 4 hour outage is not going to make any captive generation capacity go away. But, at least for the back-up generators, it will indeed reduce the time of operation of the generators by 2 hours, i.e., 2 hours less of burning diesel and spewing smoke into the air. And those industrial units that use more expensive captive power 24 hours a day will be tempted to convert to back-up systems simply from a profit and price-competitiveness point of view. Also maybe industrialists who are planning new units will factor in the 2 hours extra power supply in their planning and decide not to go in for 24-hour captive generation, but rather back-up generator systems. As you say, that's an improvement. I do feel that back-up generation capacity in India will continue for a long time, at least as in insurance policy, but I think the scope does exist to reduce the time of operation of these backup generators by building new power plants and improving the electricity grid.

4. I looked at the blog post you cited (, which argues that it will be cost competitive to build solar thermal power plants in India rather than the Tata Ultra Mega coal power plant. The basis of the argument is that CDM or carbon credits under the Kyoto framework will make the otherwise expensive solar thermal plans as cheap as the planned coal power plant. Unfortunately, I have never studied the CDM in any detail, so I don't really understand how it operates. But at least reading the blog, I think the argument does not pass the sanity check. Per the logic in that post, the Tata Ultra Mega plant itself will be eligible for $60million a year in CDM payments, since it will be eligible for $15 per million ton of CO2 averted every year compared to a baseline of 29.7 million tons of CO2. So in the post, the CDM difference between the coal plant and solar thermal will actually be $385million rather than $445million. On a broader level, I find this whole logic to be wierd. Per reports, China opens some 3 or 4 new coal power plants every month. So if these plants have a CO2 emission rate marginally below 29.7 million tons, they will be eligible for CDM credits. So China is bound to monopolize the lion's share of CDM payments simply by virtue of the fact that it builds more power plants (of whatever type) than any other country. How many billions of $$ of CDM will the Three Gorges Dam Hydroelectric plant get? This does not seem right to me. So either the post misrepresents or oversimplifies the CDM payments system, or the whole CDM thing is an incredibly stupid mechanism. In any case, Tata Power is financially hard-nosed company. If there is clear-cut commercial case to be made for large profits by building solar thermal power plants, you bet they will do it.

5. I agree completely with you that there is great scope in India for thermal power generation. Large parts of India are very sunny. But in my opinion, subsidies for these should come from the U.S. or the E.U., not from the Indian govt. For the Indian govt., I think money should first be spent on basic public services such as education, healthcare, a basic safety net, and infrastructure development, including, yes, an improved electricity grid. Nor do I think that the World Bank, whose funds are earmarked for development projects around the world, should pay for these subsidies. But I think, more important than subsidies for building solar thermal plants in India, the U.S. should probably first invest in further developing the technology of solar thermal plants in the U.S. itself. If it becomes cheap enough, companies like Tata Power will jump on the solar thermal bandwagon on their own. It might interest you to know that solar thermal energy is widely used in India. Solar thermal water heaters are fairly common throughout India. I recall that when I was a kid, we had a solar cooker at home, which I think the govt or some agency was distributing almost for free. It made excellent chicken curry too. But it was not convenient. We had to carry it all the way to the roof of our apartment building to expose it to the sun, and it took the whole day to cook. Other traditional forms of solar thermal power - drying laundry, drying various farm products, etc. are of course in widespread usage across India.

By Sid Shome on 2008 04 26

Thank you for your thoughtful response to my post on Grist. I thought I would respond here rather than on Grist, where I would have to keep things short.

Let me restate in more detail my conclusion about Tata Mundra: I believe it to be exceedingly unlikely that the addition of this supply will spur (a) new electricity connections or (b) meaningfully increase supply for households with unreliable electricity. I think it is also very unlikely that Tata Mundra will be the magic bullet that stems the rapid growth in captive generation in Gujurat and other states it supplies. So let me explain in more detail the alternative approach that I would suggest for overcoming the challenges of the India grid in a way that is climate friendly. In a system where many people are paying right now for expensive diesel electricity, it seems to me it actually makes it easier to insert renewables into the picture.

I'd like to delve into this question of captive generation, and as a subset of that the back-up diesel generators you mention. Sid, I have exactly the same conception of captive/backup generation as you do. You didn�t even mention the most critical aspect of it, which is keeping internet servers running no matter what, because when Indian back offices lose reliability they lose everything. And having idle factories is not so cheap either. So this is not simply the cost of people having to take the stairs, it is massive and worth the extra costs for electricity that industrial facilities shell out.

It�s important to note that many facilities do not just have backup generators, they have their own power plants. Except for the very largest, their power is more expensive than grid electricity. (Gujurat captive generation info is here:

When you have power plant power + unreliable grid +end-user equipment to protect against voltage fluctuations + backup generator, you have got to factor in the cost of the whole system to get the price of delivered reliable power. If you figure power plant +unreliable grid + small captive power plant + offset power used by agricultural consumers paying far below cost of supply (because now the industrial facility isn�t using ANY grid power), the capital costs of delivered reliable power have got to be astronomical. I think that�s the cost you need to be comparing to renewables (a good comparison of Tata Mundra with concentrated solar thermal power is here:

The suggestion that 4,000 MW of Tata Mundra power will offset 4,000 MW of captive generation power doesn�t take into account how captive generation decisions get made, apart from how, while I am not an electrical engineer, adding supply to a grid is not like filling a line of buckets with a hose. If Tata Mundra reduces Gujurat�s peak power shortfall from 12% to 6%, let�s be simplistic and say that cuts power outages in half for industrial facilities. A two hour power outage rather than a four hour power outage is not going to make captive generation go away. It�s conceivable that a different level of expected plant capacity utilization will make companies switch from full captive generation to backup generators. That�s an improvement, but I would still make the bet that when you add up the capital AND generating costs for both systems, together they�re more than renewables would cost.

As for the grid � I would not take the position that India should not build capacity until the grid is improved, even though in a lot of cases that�s probably the cheapest way to deliver more power to consumers, and demonstrated attention to grid maintenance has been shown to reduce theft as well. Several people have responded to me on the question of electricity theft. The question of electricity being used for amenities by poor households who can�t afford it otherwise is a tough question. But it�s impossible to get away from the fact that in the long run, the only thing that can sustain electricity use for amenities is a functioning power system that delivers reliable power that can be put to income-generating use and which is paid for so as to allow continued supply. Lifeline tariffs for poor households are a standard feature in many power sectors and it is really not electricity use by poor people who need it but can�t afford it that is pulling down the system.

As far as technical losses, even where rural electric co-ops have taken over for state utilities and brought theft down to almost zero in rural areas, the grid still loses 15-20% in technical losses. So that might be taken as a baseline of technical losses and that�s double what it could be. If Tata Mundra power is going to 5 states as the IFC says, it is going to be going a long way on a bad grid.

The real key here that I think holds promise is the fact that if you build 100+ MW of renewable power � maybe solar thermal as is being done in the U.S. with comparable insolation levels to Gujurat - and you put it right next to 100+ MW of demand instead of way off on the coast where a mega coal plant can get the water it needs, you can probably deliver a much greater percentage of it, you can improve the grid right there at much less cost to make that supply reliable, it strikes me that if utilities take a distributed generation approach � which is exactly what captive generation is � you can not only produce the utility power cleanly, you can obviate the need for backup and captive power at the scale that it exists in India. So again, line the full system cost of Tata Mundra up with backup generators and compare that to solar thermal and see how the costs come out.

On that basis, I would bet that renewables can get close if not even. But even if it�s merely close, it strikes me that that has be what public financing should be flowing to, not to marginal efficiency gains of efficient coal over less efficient coal, or even coal over diesel generators.

By Nathan on 2008 04 24

Teryn, you are absolutely right. As you very rightly point out, the whole business of carbon credits and CDM and the whole Kyoto framework is deeply flawed. The only way to halt the rapid growth of new coal power plants in India (and even more so in China) is to develop new energy technologies that are clean and cheap. This is unlikely to happen unless there is massive investment in these new technologies. Unfortunately, many environmentalists are fixated on a different idea: how to reduce energy consumption. But energy consumption is not all bad. In many parts of the developing world where people are living in poverty, increase in energy consumption would be good (less poverty in the world is good). The real question is: how to reduce emissions even while increasing energy consumption.

On a different note, an interesting article in today's NYT on new coal power plants in Europe

By Sid Shome on 2008 04 23

A flaw in Nathan's analysis is his assertion that "investing in coal generation... has the effect of spurring the construction of small-scale fossil fuel generation." As Sid points out, new electricity in any form -- whether it be from a new coal plant or efficiency gains -- should reduce the use of these backup generators.

The question is how large the cost difference is between investing in the new coal plant versus electrical infrastructure efficiency gains. AZ seems right that more analysis should be done to determine the trade-off.

If the irreliability issue is actually just one of supply and demand, as Sid says, it would seem plausible that a central station plant could address it just as effectively as investments in transmission system upgrades or in distributed generation. If the short-term cost premium for efficiency upgrades is outside the realm of Indian's economic and political feasibility, then it seems there's a good case for the plant to be built.

Nevertheless, something feels strange about having a post on our blog in support of developing an ultra-mega coal plant. I think a simple reframing of this argument would help. The larger point is that the World Bank's support of such a project illustrates 1) the CDM has serious problems, 2) massive numbers of coal plants will continue to be built without major advances in clean energy, and 3) developing nations will become sustainable to the extent that we invest in their development.

By Teryn Norris on 2008 04 23

AZ, say that you drive a 15 year old gas-guzzling, CO2 spewing SUV that is about to quit working. You're thinking of buying a replacement. Should you buy the hybrid Toyota Prius now? Or should you make temporary repairs to your SUV and wait for plug-in hybrids to become available in, say, three years? Or should somehow keep your SUV running and wait till solar-powered cars become available in, say, 15 years? We all have to make trade-offs between the long term and the short term, balancing various factors. Most reasonable people would agree that considering the expected increase in electricity demand in India, the status quo, with frequent power cuts, and proliferation of extremely dirty small-scale electricity generators, is simply unacceptable. You do the best that you can with the technology available today. Hence the coal-fired power plant. The trade-off seems eminently reasonable to me. I am sure the professionals at Tata Power, and the World Bank and other banks that are financing the project have done their own detailed quantitative trade-off analysis as well. Unfortunately, clean renewables (except hydro) have not yet reached the point of being commercially viable on a large scale. This is exactly why Breakthrough is pushing for large public investments in these relatively nascent technologies. Hopefully, such investments will indeed be made and new, cheap, clean energy technologies will be developed that will be able to hold their own in trade-off analyses with coal-fired power plants in the future. I also want to point out that it may be possible, in the future, to retrofit this and other coal power plants with carbon-capture-and-storage technology that can greatly reduce carbon emissions (see ). So even if the plant remains in use for a long time, emissions could potentially be reduced.

By Sid Shome on 2008 04 23

Interesting post. Essentially the argument is that large coal plants are less dirty than the alternative, which is portable diesel power generators. This is relatively convincing, except I think it misses one point. I am certainly no expert on the Indian energy sector, but it seems to me that building such a large coal plant assumes that it will be in use for a long time. Regardless of the advances in technology, which Breakthrough advises is our only way to mitigate global warming impacts, this coal plant will remain a relatively dirty form of energy production until it comes off-line. Backup diesel generators, however, have much lower fixed costs and are, if Mr. Shome is describing their intended use correctly, not meant as a permanent fixture of India's power supply portfolio. Therefore, there is a tradeoff between short term, marginally cleaner energy and long-term barriers to a clean Indian power grid. This is not to say that the trade-off isn't worth it, but some analysis would need to be done to make that determination. Just because it is cleaner than the status quo does not immediately suggest that it should be built.

By AZ on 2008 04 22