How much would it cost to retire a coal-fired power plant?

David Fogarty, reporting for the Straits Times:

SINGAPORE – Japanese conglomerate Mitsubishi Corporation on May 7 joined Singapore’s Keppel and investment platform GenZero in a climate initiative that aims to retire a Philippine coal plant early using funds from a new type of carbon credit backed by the Monetary Authority of Singapore (MAS). 

[…]

The aim is to retire a 246 megawatt (MW) coal plant in Batangas province, South Luzon, by 2030 – ahead of its scheduled closure in 2040 – and replace it with renewable energy and battery storage.

Shutting it down a decade early could save 19 million tonnes of planet-warming carbon dioxide emissions – and cut local air pollution.

The catch is that closing the plant early is costly. Replacing it with green power generation and new power lines, as well as compensation for the lost 10 years of electricity income, would cost over US$1 billion (S$1.29 billion), said Mr Eric Francia, president and chief executive of the plant’s owner Acen, which is the listed energy platform of Ayala Group.

“We’re still finalising the numbers, we’re talking about US$1.5 billion-plus overall cost,” he told The Straits Times at the GenZero Climate Summit 2025, being held from May 5 to 8 at the Sands Expo and Convention Centre.

Fully replacing the coal plant with the same level of on-demand power would require 1,000MW of solar, 250MW of wind, and 1,000MW of battery energy storage, according to The Rockefeller Foundation, which is also involved in the project.

US$1.5 billion seems high.

For comparison, in a 2023 report, MAS and McKinsey modelled what it would take to shorten the lifespan of a hypothetical 1 GW coal plant in Indonesia by 5 years. At a 50% capacity factor and 8-9% discount rate, retiring the 1 GW plant early will only cost USD 70 million. This number does not include investments in and revenues from renewables that would replace the retired coal capacity. The early retirement would cut 23 million tonnes of CO₂ emissions. If these avoided emissions were to be sold as carbon credits to close the 70 million dollar cost gap, each tonne of CO₂ would cost US$11-12.

Now, ACEN is asking for US$1.5 billion to retire a coal plant that’s one quarter the size (246 MW vs 1 GW). Sure, the Batangas plant will shorten its lifespan by 10 years instead of 5. Sure, it may be operating at a higher capacity factor, generating more future revenues and carrying a higher NPV. This price tag includes investments in renewables and storage to replace the retired coal plant, but shouldn’t you also account for future revenues generated by the renewables?

Speaking of renewables, I don’t believe you need 1 GW of solar, 250 MW of wind and 1 GW of storage to replace a 246 MW coal plant. The cost of adding intermittent renewables and the need for storage both depend on how much intermittent renewables is already on the grid. As of May 2024, the Luzon grid had a total installed capacity of 20 GW, with only less than 8% intermittent sources (1.2 GW solar and 337 MW wind) and 363 MW storage. Would you believe them when they say they need to almost double their solar and wind capacity, and quadruple storage, just to retire 1.6% of the fossil fuel?

[Francia] said he hopes the Singapore Government might be among buyers of the credits and linked the future price of the credits to the Republic’s estimated carbon tax price of $50 to $80 per tonne of emissions in the coming years.

“We’re anchoring this to the Singapore carbon tax. Hopefully, we would be closer to the lower end of the $50 to $80 range,” he said.

Compared to the US$11-12 per tonne estimated for the hypothetical Indonesian plant, what ACEN is asking for feels more like a shakedown.