MSEDCL’s new coal + solar tender was Rs 4.08/kWh. How much would wind + storage + solar cost for the same profile?

This week, MSEDCL announced that it had awarded a 1600 MW coal + 5000 MW solar tender at Rs 4.08/kWh. The combination of coal + solar is odd, as due to it’s slow ramping speed and high cost of turning off and on during the day, coal is uniquely poorly synergized with solar. This is in contrast to storage which can absorb extra solar during the day and ramp up to provide firm output during evening/morning peaks. Even hydro or gas have a better synergy with solar as they can be ramped up and down as solar goes down and up.

In short a coal + solar profile *cannot* be compared with “RTC” or “Firm, Dispatchable” Power. So it is worth asking, how much would it cost wind + solar + storage to meet the same 1600 MW¹ firm + 5000 MW solar profile. Figure 1 summarizes the results. In short, this coal + solar combination costs Rs. 1.4/kWh more than solar + wind + storage for the same profile, at a cost to MSEDCL and its consumers of ~2800 Crore *per year*.

Figure 1: Comparison of coal + solar vs. wind + storage + solar

To start with, the Rs. 4.08/kWh is a “blended” price, with the solar at Rs. 2.7/kWh, and the coal with an assumed PLF of 85%. This works out to a cost of the coal plant of ~Rs 5.4/kWh. However, this assumes a “flat” profile-trying to match the load profile to coal, rather than the other way around. In reality, load fluctuates up and down. While the tender correctly notes that agricultural load is “very price sensitive” and therefore can be shifted to the day so it can be met by solar, the coal must still ramp up to meet the daily peak (usually evening) and ramp down during periods of lower load (eg. late night/early morning), as shown in Figure 2.

Figure 2: Typical Load profile for 1600 MW thermal + 5000 MW solar

I discussed the extra cost due to lower heat rate and higher fixed cost per unit here, but the point is the coal “as run”, will be more expensive than as assumed at 85% PLF. This gap will continue to widen as India develops, with greater air conditioning use and higher reliability expectations driving “spikier” load profiles which coal is particularly poorly suited to. Currently, load spikes are clipped using load shedding to help maintain coal plant PLFs, but as India becomes a Viksit Bharat, this will no longer be tolerated.

In contrast to meet the same load profile as 5000 MW solar + 1600 MW “flat” thermal would require 7400 MW solar, 2800 MW wind and 1500 MW/4800 MWh of battery storage, at a cost of ~Rs 3.32/kWh, about 76 paisa/kWh cheaper than as quoted for coal + solar for the same load profile, with the same level of availability (85%. Though it is worth noting that unlike RE + storage, coal is typically exempt from penalties when it underperforms).

Figure 3: Solar + Wind + Storage to meet 1600 MW firm + 5000 MW solar, on a typical day

As shown in Figure 3, storage can ramp rapidly both to meet peaks as well as integrate fluctuating wind. If we consider Maharastra’s actual load profile instead of a “flat” load profile, the cost is even lower (Rs 2.84/kWh). Because load is lower at night/early morning, we need slightly less wind and storage to meet the same availability requirement. This is a crucial distinction. If we want to meet actual load profiles instead of “flat” load, coal is *more* expensive (as it costs more to run at part load), whereas RE + storage is *less* expensive (as we do not need as much wind + storage overbuild).

And this is the best case scenario for coal. While Rs 4.08/unit is the suggested tariff, coal prices are “indexed” and passed through to consumers. Given that India imports about 21% of it’s thermal coal, this leaves Maharastra consumers exposed to commodity price risk. Furthermore, CERC has mandated that coal plants are allowed a 5.5%/year O&M escalation. In contrast, RE + storage costs will continue to decline. Finally, I often here that coal is required because “we can’t build RE fast enough”. But new coal will take at least 48 months to build (I just heard 72 months for one plant), whereas wind and storage can be built in 24 months² meaning wind + solar + storage can meet Maharastra’s growing power demand much faster than solar + coal. And of course this ignores the other benefits such as lower air pollution, more jobs, higher government tax rates, no water dependency, etc.

MSEDCL consumers continue to face step tariff hikes due to over-reliance on expensive coal plants, such as the recently built Gadarwara (Rs 6.18/kWh), APML (Rs 9.19-10.65/kWh) and Parli (Rs. 6.58/kWh). Even if this coal plant manages to meet it’s optimistic cost assumptions, it will still be costing MSEDCL consumers at least 2800 Crore per year more than the worst case for wind + solar + storage. It is time Maharastra regulators started standing up for consumers and try to minimize cost and reliability. This tender should be scrapped and replaced with a technology-neutral tender for 1600 MW firm power (and a separate 5000 MW vanilla solar, if MSEDCL wishes to do so to meet agricultural load).

If the supporters of coal genuinely believe it is economic, let it compete on an even playing field with wind, solar and storage.

1. The thermal capacity is 1600 MW, but due to losses at coal plants themselves, the firm electrical output is 1470 MW ↩︎
2. I have heard from many developers there are concerns around allocation of land/connectivity, particularly for wind, delaying projects to 30-36 months. This is all the more reason for states to prioritize their precious time on supporting land/transmission connectivity for wind, but even in the worst case, this is still much faster than the best case for new coal ↩︎