Solar Could Beat Coal to Become the Cheapest Power on Earth In Less Than a Decade (bloomberg.com)

Solar power is now cheaper than coal in some parts of the world. In less than a decade, it's likely to be the lowest-cost option almost everywhere, reports Bloomberg. From the article: In 2016, countries from Chile to the United Arab Emirates broke records with deals to generate electricity from sunshine for less than 3 cents a kilowatt-hour, half the average global cost of coal power. Now, Saudi Arabia, Jordan and Mexico are planning auctions and tenders for this year, aiming to drop prices even further. Taking advantage: Companies such as Italy's Enel SpA and Dublin's Mainstream Renewable Power, who gained experienced in Europe and now seek new markets abroad as subsidies dry up at home. Since 2009, solar prices are down 62 percent, with every part of the supply chain trimming costs. That's help cut risk premiums on bank loans, and pushed manufacturing capacity to record levels. By 2025, solar may be cheaper than using coal on average globally, according to Bloomberg New Energy Finance. The solar supply chain is experiencing "a Wal-Mart effect" from higher volumes and lower margins, according to Sami Khoreibi, founder and chief executive officer of Enviromena Power Systems. The speed at which the price of solar will drop below coal varies in each country. Places that import coal or tax polluters with a carbon price, such as Europe and Brazil, will see a crossover in the 2020s, if not before. Countries with large domestic coal reserves such as India and China will probably take longer.

Comments

Re:What type of solar

[Maxo-Texas]

I suspect as solar becomes ubiquitous we may see more DC options.

Photo voltaic has become very compelling plus we don't fund people who want to kill us when we buy photo voltaic so that's always a plus.

But molten salt is pretty compelling for solar as well.

Coal is already uneconomical compared to other resources even without considering the pollution cleanup costs. Old coal plants didn't have to comply to the new pollution laws until last year (well 2015 so I guess now barely two years ago) and were polluting large areas with mercury.

Nuclear is great as long as you ignore decommissioning and fuel storage and human nature. i.e. humans get sloppier and cut more and more corners over time until something bad happens. I'd feel more comfortable if nuclear were restricted to small (5000 house) self contained plants which didn't even allow humans in the loop and which shut themselves down automatically. And we need to build a breeder reactor to reduce the volume of nuclear waste by 2 orders of magnitude. But it has to be crazy secure. As in put it on an army base secure.

Solar, wind, and tides are the way to go tho. All have minimal cleanup costs, minimal problems on failure, fail by tiny pieces rather than as a whole, and costs are plummeting.

Re:What about at night?

[Rei]

No. But it's not prohibitively expensive, generally adding a couple cents per kWh to your total costs**. The amount of peaking/storage required depends on a lot of factors, including climate, diversity of generation (e.g. wind + solar has much higher statistical reliability than just wind or solar, as they tend to run counter to each other), and the amount of long distance transmission (HVDC/HVAC), for 1) geographic diversity of weather, 2) sharing common peaking resources, and 3) timeshifting of loads/generation. A recent study in nature estimates that a nationwide US HVDC network would cost 0,3 cents per kWh but save 1,1 cents per kWh in generation/peaking hardware costs. The cost of peaking (and type) depends on location. Hydroelectric turbine house uprating makes for very cheap peaking where available (transforming baseload hydro into peaking hydro). Pumped hydro can be affordable, but only in limited areas. Batteries are marginal at present, but are likely to become highly competitive over the next decade. In the US, where natural gas is cheap and plentiful, the vast majority of new peaking capacity is NG. In countries where natural gas is expensive, other fossil fuels are used.

Also note that up to a certain level of penetration, solar actually does more to help remove variable generation (load following plants) than it imposes (peaking), as daytime loads are higher than night, and are higher on sunny days than cloudy days.

** - A peaker that's used only several hours a year may charge $2/kWh or so... but you're not buying a lot of kWh from it. A load following plant that's used a bunch every day may only charge $0,1/kWh... but you're buying a lot of kWh from it. It all depends on what sort of power you're needing to buy.

Re:What about at night?

[Rei]

The odds of the weather across all of North America being cloudy are virtually nonexistent. Power systems don't ever guarantee 100% uptime (not today, not in the future), because that involves planning against events that are finite but absurdly improbable. You plan for whatever 99,9+% uptime targets that you deem appropriate, with plans for how to fail gracefully. A single front does not stretch the entire width and height of North America. But nonetheless, diversity in power sources is good. Low pressure systems tend to bring clouds, but they also bring wind - just like how wind peaks at night, in contrast to solar's daytime peak. Also, peak solar seasons vary a bit from region to region, while wind seasonal peaks vary greatly from region to region (in the US, the west coast has a summer peak, while the central and eastern US have a winter peak).

Don't take my word that you can achieve statistically significant uptime without unrealistic peaking costs - read any of the studies on the subject. There've been a lot of them. It requires no new tech and no storage - although those have the potential to make things even cheaper and easier.

Beyond peaking and storage, you have entire industries where their costs are predominantly driven by electricity costs. Such industries are often quite willing to engage in curtailment agreements with power companies in exchange for cheaper rates. Which basically doubles as peaking. Here in Iceland, for example, we have aluminum and silicon smelters that import all of their raw materials, and export almost all of their products; it's worth it to ship everything to and from a remote island just for the cheap power. And boy do they gobble it up - even the smallest of the aluminum smelters uses more power than all homes and businesses combined. And beyond time-shifting of entire industries, there's timeshifting for particular hardware units in other industries. For example, chillers rarely run 24/7, and can also be timeshifted.

I'll reiterate that I think diversity is important. Picture a 100% solar world in which you have even intercontinental power transmission, ultra-high voltage DC doing hops of thousands of kilometers at a time. All of North America interconnected, running into Siberia and China from Alaska, to South America through Central America, and to Europe through Greenland and then Iceland (where there's already a lot of prep work underway for power lines to the UK). You have the whole planet equallizing you out and timeshifting - virtually no peaking/storage at all required. All well and good!.... until a major volcano goes off. When Laki here went off in 1783, the huge quantities of gases it kicked out altered the global weather so much that the Mississippi River froze at New Orleans. Not from clouds, but a global stratospheric haze layer. There was plenty of wind that year, mind you, but very little sun! Being single-source dependent leaves you vulnerable. Regardless of what that source is.