I’ve seen a lot of debates about this, and I want a factual analysis. Assuming that there are no subsidies or tax breaks, which sources are really the cheapest per kWh?
I’d like to see how nuclear, coal and natural gas really compare to hydro, solar, wind, and biomass/Waste-to-energy. Many of these are highly subsidized, and I am wondering what the real price to consumers would be if all subsidies are removed.
Do you want to include just the end user dollar cost or also estimates of externalities?
A useful metric here (not, by any means, the only one) is “lifetime cost of electricity”. This aims to measure the unit cost (e.g. per MWh) by adding the operating costs to the discounted capital costs over the lifetime of the asset.
This allows comparison between technologies such as wind and hydro, on the one hand, with low operating cost but a high capital cost (per unit generated) and sources with a high operating cost but where the capital cost per unit generated is much lower, such as gas turbines.
Unfortunately it’s still not straightforward as the result can be highly dependent on the discount rate you assume, as well as how you treat financing costs (which are a major component of the cost of developing some kinds of assets, most notably nuclear). You also have to be clear on whether you want to consider the commodity prices as they are now, or a projected forward price curve.
Finally, there is considerable debate as to what constitutes a subsidy. Do you consider externalities such as pollution to be a form of subsidy? And if so, can you get consensus on how to value them?
Edit: Another point I forgot to mention is that you may choose to include the “full system costs” - not just the local cost of building and operating the wind farm but the cost of transmission network reinforcement, backup generation and reserve.
End user dollar cost. I doubt there is a factual GQ answer to the externalities question.
Yeah, but no. If you are really serious about figuring out “real costs” of just about anything as pervasive as energy production, you are going to have to make a lot of assumptions, and ignore a lot of real factors…OR accept that it’s a hell of a lot more complicated than almost all people debating it are willing to recognize.
Very brilliant insight above, to ask how you calculate the “tax break” of allowing pollution. Another sneaky “subsidy,” is all the arranged local monopolies.
But I want to point out another entire aspect of figuring “cost,” which almost everyone overlooks. And that is, that the entire structure of the assignment of values, itself. A fair amount of value assignment, is altogether arbitrary, often being based more in tradition, than in practicality.
I keep thinking about how we USED to think about things like mining, as compared to how we think of it now. For millennia, most societies thought that the only cost in mining, was the cost of keeping miners alive long enough to get the stuff being mined. But then we learned that there was a huge cost being ignored, to how the environment of the mines, and of the areas where the mine products were used, changed how people had to live.
Not just the concern of pollution, but also the cost to the society, of making themselves dependent on a single resource, or even just the subtle cost, of allowing reliance on an energy source, to force us to adjust our entire society to cater to the resource.
Heck, some things that are thought of as “job-and-wealth producing industries,” could also be seen as “society shaping addictions.”
By the way, just a sort of related thought experiment: we once thought that water resources were limitless. That we could throw anything into the river or the sea, and it would just go away. But we eventually found out that there IS such a thing as completely changing and even destroying a water resource, by dumping too much “stuff” into it, and we can certainly set up too many damns and channel too much of it into crops, trying to utilize it.
What if we switch to wind power, and build so many wind farms, that we actually “use up” the wind itself, and change the world down wind of the turbines?
I suspect that in the long term, wind, solar and hydro are cheaper. No continuous fuel costs. But each has initial installation, maintenance and replacement costs. Exclude the fuel source and most electric generating systems, except PV panels, have the same devices to create electricity. Rotating generators. Combustion and hydro have fewer but more intense generators.
A dam is a huge initial construction cost. But has a very efficient, concentrated generating facility, that is very self contained.
A combustion facility is quite cheap to install and concentrated. But requires the large fossil fuel industry infrastructure to run. That industry cost is very hard to pin down. The costs seep into a lot of things. It is a complex industry that can impinge electricity production.
Wind and solar PV are expensive to install, but no fuel costs. Self contained. Solar PV is very easy to calculate maintenance costs. Wind is more variable. Though extra wind maintenance cost may relate to extra power produced at profit and be positive or neutral.
It is very difficult to get real numbers on most systems. There are subsidies, hidden costs, that are separate from the reality of the true costs.
But looking at the mechanics is a fairly good way to estimate. As long as you include all the mechanics.
My opinion and the mechanics? Solar. Direct solar to DC panels. Not as the immediate best cost solution. But the ultimate best solution. It is the least complex system. It is still increasing in efficiency. It can be installed all over the place. Open land, on rooftops in a city.
A wild card, is viable superconductive materials. A super conductive power grid would be a massive boost to decentralized power production. Sadly, it seems super conduction is lagging solar and battery progress.
To arrive at a “true” cost one would have to go right back to the costs related to sourcing all the components (including environmental) through production and distribution to decommissioning. This last is a major factor in the cost of nuclear, but those windmills will come down eventually.
A major drawback of solar is that it functions, at best, about half of each day. It doesn’t work nights or on cloudy days and people expect electricity to be available 24/7/365. So you have to install redundant non-solar systems to provide electricity on demand or you have to have massive electrical storage systems. That adds significant costs and complexity.
Expanding on what Alley Dweller said, wind and solar both suck because they aren’t consistent, and there isn’t much energy storage in most power systems. Many power systems are forced to run other types of power plants either at idle or at reduced capacity so that they can quickly increase their power and make up for inconsistencies in solar and wind generator. This wastes a lot of energy and resources in those idling and under-utilized plants.
If you are trying to figure out the true costs of these systems, you also need to factor in the wasted energy in other plants so that the system overall can provide consistent power to its customers, rain or shine, night or day, and regardless of the current wind conditions.
As energy storage gets cheaper, solar might become much more attractive, but right now, it’s god-awful expensive if you want to have it capable of generating its own power without needing some fossil or nuke plant backing it up.
There are ways of generating energy from solar power that work 24/7/365, but they aren’t things that you can just shove on a rooftop somewhere. You can take a whole bunch of mirrors and focus them on a container of salt, which will focus so much of the sun’s energy there that the salt will melt. You then use the molten salt to heat steam, and use that to drive a conventional steam turbine generator. With a large enough system, the salt will retain the heat through the night, allowing the plant to continually generate power even when the sun isn’t shining. These types of plants work well in sunny areas like the southwestern U.S. but if you live in an area where the sunlight isn’t quite so consistent, you can’t keep the salt molten and that type of plant doesn’t work so well.
A lot of these molten salt plants are new and are likely subsidized as they are still somewhat experimental. I don’t have any data on their costs or subsidies though.
Wind and solar are also sometimes used to do things like pump a bunch of water uphill into a reservoir. At night or at other times when additional energy is needed, that water can be channeled down through conventional hydroelectric generators to recover that energy and make up for the wind/solar inconsistencies. Obviously, all of this adds a significant amount of extra complexity and cost.
DC (direct current) also sucks. There’s a reason (actually a bunch of them) why Edison lost to Westinghouse and Tesla’s AC. In order to be practical, any solar power system that generates DC is going to also need an inverter to convert the DC to AC.
Actually DC is more efficient for longer distance transmission and for certain other situations (e.g. underwater cables)
Brian
As I understand it, under almost all measures, hydro is the cheapest by a considerable margin. It also scores well on a variety of other measures, such as consistency of availability (at least in the short term) and speed with which it can be ramped up or down to meet demand. The main problem with hydro is that it’s already very close to saturated: All the good places to build dams already have dams built at them. The main way to increase the proportion of our power that comes from hydro is to decrease our consumption.
And AC and DC both have their benefits and drawbacks. Given the technology of the early 1900s, the benefits of AC were far more significant than those of DC, and so we built up all of our infrastructure based on AC. As technology has advanced, though, that has changed. If we were rebuilding the grid completely from scratch now, but with modern technology available, we’d be fools to use AC, as current tech can almost completely eliminate the drawbacks of DC. The only reason we’re still using AC is inertia from the large amount of equipment we already have in place that’s based on AC.
DC does have benefits. AC lines have to be designed to handle their peak voltage, but the effective power you get out of them scales to the RMS value, not the peak value. DC always runs at peak, so for the same wire size, insulation standoffs, etc. you can push more power through with DC than you can with AC.
DC also doesn’t suffer from capacitive and inductive losses in the transmission line. Reactive losses get very significant in undersea cables, so DC definitely has an advantage there.
But, DC has its drawbacks as well. An AC transformer is just two coils of wire around a hunk of iron. You can’t get much simpler than that. DC transformers do exist, but they are much more complicated, and expensive.
The switchgear for DC is also more complicated and more expensive. When you open an AC switch, it tends to automatically extinguish the arc simply due to the fact that the AC voltage drops to zero twice during the AC cycle. DC remains at a constant voltage, and won’t naturally extinguish the arc, meaning that the switch gear has to be designed with arc suppression.
The transformer and switchgear issues mean that DC is only better once you get out to a long enough distance that the wire savings is greater than the cost of the transformers and switchgear at either end of the DC line.
The transformer/switchgear costs mean that DC will never be superior to AC for electrical distribution, at least not in the foreseeable future.
Kedikat’s comment about DC referred to using solar on rooftops, which means the DC is only being used in the local building, or at most for local power distribution over a relatively short distance. In that context, DC always sucks.
The transformer and switchgear issues are why Edison lost and why we use AC today.
ETA: Chronos’s post wasn’t there when I started typing. If we were to redesign the electrical grid today, using modern technology, we would certainly use a lot more DC on the transmission side of things. But not on the local distribution side.
Good point. Note also that the calculated cost of coal “production” ignores that after you’ve mined it, the coal isn’t there anymore!
Anyway, a factual GQ answer for OP is that land-based wind power is the cheapest electricity … at least for places where there is harvestable wind. Similarly, hydro power is cheap where harvestable. Other renewable sources are more expensive at present and, again, dependent on availability.
Carbon-based fuels are cheaper than nuclear if you don’t mind global warming. Otherwise state-of-the-art nuclear is much cheaper than carbon, though some worry about risks like Chernobyl…
Someday soon, solar and/or fusion will be cheaper than carbon or nuclear! 
How do you calculate the hidden value of the US Navy keeping a carrier group near the Straits of Hormuz now and then to keep the oil tankers safe besides having an occasional shooting war. Also those people plus other navies that keep piracy low on the open seas for the same tankers. Those costs needed to be included with the petrofuels. Along with the cost of the pollution which isn’t fully realized. The health issues of wind power have not been acknowledged, documented, and assessed on a “generally accepted” level yet either. Where do the materials for the solar panels come from, how is it acquired, and how much of that cost is dumped on society at large? Seems the bigger the net, the more each individual energy source has hidden subsidies that are borne by society.
The traditional coal plant killed a huge number of people because of the air pollution it produced. Even now:
But the OP isn’t concerned about externalities. So the cost of any coal plant actually being built could be much lower if it weren’t concerned with lowering pollutants (not even considering the global warming issue).
I agree there’s no factual answer for externality cost and it can devolve into a plug number to make renewable sources look more/less competitive than they actually are according to opinions or preconceptions. Also a lot of externality ‘cost’ is really risk. It’s rational to pay to avoid risk* but it’s different concept than a fairly cut and dried cost of a powerplant, even with the uncertainties of future capital and fuel costs. Meaning mainly climate implication. I think the cost of countering ‘conventional’ pollution can be to some degree considered differently, as more of a given for eg. what equipment a coal plant needs that a gas plant does not.
The other complication related to renewables is availability. This doesn’t factor as much into comparing the cost of coal and gas fired plants etc. But it might be that a grid largely powered by solar or wind has to have a lot of standby conventional capacity to get through dark/cloudy/calm periods. For awhile when it’s a small % of capacity it can just take advantage of older ‘paid for’ fossil plants which get used less. But in the longer run/larger scale it’s too optimistic not to take into account eventually building new standby fossil plants, or renewable storage facilities (former might be cheaper). Or if an energy source requires a lot of grid work (long distance DC etc) that’s part of the cost also.
OTOH for sources which work on a distributed basis like local solar panels (though also small gas fired congeneration plants) it’s the other way around. You have to credit those with the cost reduction of less building/maintenance of grid at the margin, all else equal.
But in general the implication that renewables can replace fossil for a large % of use without much higher end user cost (not counting externalities), at today’s technology, is too optimistic when all aspects except climate externality are considered. So that gets back to cost/risk of that externality, a preference issue.
*eg. accept a lower expected return on treasury bonds than stocks, rational investors do it everyday at least for some portion of their holdings. On the basis of ‘cost’ not including risk, stocks are just better and you’d never have bonds. Which some people believe. Converting risk to cost, or return you’re willing to forego, depends on personal risk tolerance.
If the worst case models are really correct, I think “preference” might be a bit too weak of a term. The equator would be uninhabitable to human beings who aren’t inside actively cooled habitats or space suits. Lots of key coastal real estate would have flooded, and obviously which land is arable and which land isn’t would be flipped upside down. It wouldn’t be the end of humanity - among other things, it would make certain regions of the world quite nice, like the northern regions of Canada, Greenland, parts of Antarctica possibly, and Siberia - but it would be pretty damn expensive in dollar cost. Essentially it’s a wiping out of the value of most of the land and most of the real estate on the entire planet.
But yes, I agree with you. First let’s talk about what a practical renewable grid would look like. It’s real simple. It’s wind generators, solar panels, mass produced grid storage batteries (lithium or flow), and backup generators. Solar produced power is getting buffered to the nights through the batteries - that is, there’s about enough battery capacity to make it through the average night if there’s an average amount of wind as well. The backup generators are reciprocating natural gas engines that start and run during the statistically infrequent cases when it’s a cloudy, high power draw day combined with a calm night.
The problem is if you think about it, the point of the backup generators is to save burning natural gas. Which is extremely cheap without a tax levied to compensate for the pollution caused by burning it. So since you need a lot of backup generators - probably something like half the baseload capacity (baseload = average monthly power draw for a region) - why not run the generators you had to pay for anyway longer? You need less batteries, less solar panels, it’s all good.
Following this train of logic you end up with a grid that is basically just wind generators and natural gas. If the power grid were brand new and unsubsidized - that is, if we didn’t have all these other forms of power generation or wind/solar propped up with subsidies - that would the entire grid. Just look at the numbers. Cost of electricity by source - Wikipedia
Don’t forget hydro in that mix. Even without green incentives, it’s already a significant chunk of power generation. And combined with a smart grid, smart metering, and efficient usage, it just might be able to replace the gas turbines in your proposal, or at least a significant number of them.
Aren’t they pulling down the dams ? All over the world ?
I read through the earning statement of Brookfield Renewables to try to figure out a factual answer.
This was complicated by the fact that they produce both wind and hydro. About 13GW total. The entire company received $23 Million in tax credits for renewable energy.