The Energy Collective Group

This group brings together the best thinkers on energy and climate. Join us for smart, insightful posts and conversations about where the energy industry is and where it is going.

9,876 Members

Post

Comparing the Costs of Renewable and Conventional Energy Sources

By Mitch Tobin

You can’t get too far in a discussion about the nation’s electric power sector without running into the question of costs.

How do renewable sources, such as solar and wind, stack up against fossil fuels, such as coal and natural gas?

How much will it cost utilities and ratepayers to build—and operate—a new power plant?

To illustrate how the various energy technologies compare, we’ve created a set of interactive dashboards that summarize how much it costs to generate power. The data show that utility-scale solar and wind installations are now competitive with conventional coal- and gas-fired power plants. Moreover, wind and solar costs are projected to steeply decline in the years ahead.

Before delving into the data, it’s worth noting that the number of dollars it takes to build and operate a power plant is a somewhat narrow definition of costs. It doesn’t, for example, include what economists call externalities, such as the cost of air pollution or climate change impacts. With a broader definition of costs, low-carbon technologies would perform even better than fossil fuels.

Components of levelized cost of energy

Our dashboards present data on what’s known as the levelized cost of energy. In essence, this analysis offers an apples-to-apples comparison of the costs of financing, building, operating, and maintaining a power plant. The values are expressed in dollars per megawatt-hour.

One of the most widely used levelized cost studies is conducted by Lazard, an international financial advisory and asset management firm. Their latest version of the study, version 8, was released in late 2014. The graphic below summarizes the cost components of 16 different energy technologies evaluated by Lazard: 10 of them are alternative (which includes mainly low-carbon, renewable technologies), and six are conventional (which includes fossil fuel sources and nuclear).

       Components of levelized cost of energy

Onshore wind has the lowest average levelized cost in this analysis at $59 per megawatt-hour, and utility-scale photovoltaic plants weren’t far behind at $79. By comparison, the lowest cost conventional technologies were gas combined cycle technologies, averaging $74 per megawatt-hour, and coal plants, averaging $109. These numbers are the average of Lazard’s low- and high-end estimates (see their study for more about their cost calculations).

Looking across the 16 technology types, the 10 alternative technologies cost an average $147 per megawatt-hour, $18 less than the conventional approaches. “Certain Alternative Energy generation technologies,” Lazard wrote, “are cost-competitive with conventional generational technologies under some scenarios.”

By dividing the costs among capital, fuel, and operations and maintenance (O&M), you can see some dramatic differences among the technologies. Many renewable technologies, such as wind, solar, and geothermal, may not be cheap to build, but they have no fuel costs once they’re up and running, and generally have lower O&M costs as well.

Wind and solar costs falling

The levelized cost of some wind and solar technologies has plummeted in recent years. The graphic below shows that the average cost of onshore wind has fallen from $135 per megawatt-hour in 2009 to $59 in 2014. That’s a 56 percent drop in five years. The cost of utility-scale photovoltaic technology has plunged from $359 per megawatt-hour in 2009 to $79 in 2014, a 78 percent decline. Lazard attributes these falling costs to “material declines in the pricing of system components (e.g., panels, inverters, racking, turbines, etc.), and dramatic improvements in efficiency, among other factors.”

       Levelized cost of energy: wind and solar
 

Other forms of solar power are expected to get even cheaper in the next few years. The graphic below shows that rooftop residential solar costs are expected to decline 42 percent between 2014 and 2017; for commercial and industrial photovoltaic installations, Lazard forecasts the levelized cost will drop 28 percent over the same period. “More efficient installation techniques, lower costs of capital and improved supply chains” are the driving forces behind these projected cost reductions.

       Levelized cost of energy: solar PV projections

Comparing the technologies

A variety of considerations—aside from cost—determine when, where, or how a technology is used. Although wind and solar are now cost-competitive and offer many health and environmental advantages over fossil fuels, these are still considered intermittent sources because the sun isn’t always shining and the wind isn’t always blowing). As a result, wind and solar are unable to entirely replace the services that certain conventional “baseload’” sources provide to the system. That said, utilities and power operators are discovering a multitude of new ways that low-carbon technologies can offer more value to the power grid across the board.

Transmission is another important issue, especially for power sources that are limited to certain areas, but it can also allow power plants to serve faraway customers.

Other resources

There are many other resources on the web for analyzing energy costs:

  • The U.S. Energy Information Administration’s Annual Energy Outlook 2014 provides a different set of levelized cost figures for a variety of technologies.
  • The National Renewable Energy Laboratory also offers an online calculator that lets you plug in variables, such as capital and fuel costs, to generate LCOE estimates.
  • The Department of Energy’s OpenEI offers a Transparent Cost Database that compares the results from dozens of cost studies

This post was originally published on Energy Innovation’s blog page.

Energy Innovation's picture

Thank Energy for the Post!

Energy Central contributors share their experience and insights for the benefit of other Members (like you). Please show them your appreciation by leaving a comment, 'liking' this post, or following this Member.

Recent Comments

Math Geurts's picture
Math Geurts on February 12, 2015

There is more than LCOE and external costs: integration costs.The integration costs will rise in the future, specially for solar in area’s far from the equator.

http://neon-energie.de/wp-content/uploads/2014/12/Hirth-et-al.-2015-Integration-Costs-Revisited.pdf

 

 

 

 

 

Nathan Wilson's picture
Nathan Wilson on February 12, 2015

It is great to see the levelized cost broken into components.  That makes it easier to make a realistic comparison.

For example, generally coal plants can be replaced one-for-one with nuclear plants, so their levelized costs can be compared directly.  

But to replace coal plants with solar and wind, some combination of battereries and natural gas backup would be required.  So add the fixed costs of the gas/batteries to the levelized cost of the solar/wind and compare that to the levelized cost of coal.  Notice that gas peaking plants have a much higher levelized fixed cost than gas combined cycles even though peaker plants cost slightly less to build; that’s just the effect of operating at lower capacity factor (e.g. with high wind/solar penetration).

Incidentally, you can make the solar, wind, and nuclear options look better compared to fossil fuels by choosing a low interest rate (see the Lazard report p.12, Sensativity to Cost of Capital).  Note that utilities can always borrow money cheaper than homeowners and private developers.

Wayne Lusvardi's picture
Wayne Lusvardi on February 13, 2015

Garbage in, garbage out goes the saying. 

The methodology of looking at comparative costs or renewable energy is like comparing a half truth with a whole truth. What are the full costs of renewable energy?  And moreover, which is cheaper at displacing C02, renewables or conventional energy?   Renewable energy can’t compete with conventional energy as to the net cost of displacing C02 because it is intermittent.  So the above “study” only compares the cost or renewable energy for, say, 6 hours per day for solar power and triumphally claims it is cheaper than conventional power sources.  But what is the full 24-hour cost of renewable power?  

Europe has found that the cost of natural gas unsurprisingly doubles when replaced by renewables.  So add the doubled cost of Nat Gas power for, say, 16 hours per day with the cost of renewable power for 6 to 8 hours per day and you would get closer to the real cost. 

Economist Charles Frank of the Brookings Institution has developed a way to better compare renewable energy by measuring the amount of C02 displaced and at what cost compared to conventional energy sources.  Based on that measure, renewable energy can’t compete with hydro, nat gas, or nuke power as to cost.  See Charles Frank “The Net Benefits of Low Carbon Electricity Technologies”  

Schalk Cloete's picture
Schalk Cloete on February 14, 2015

The US is probably the best country in the world for the deployment of wind/solar energy due to the combination of a number of factors: 1) excellent wind/solar resources (wind and solar capacity factors are literally double the global average), 2) lots of space 3) stringent coal regulations greatly inflating capital costs, 4) abundant natural gas for balancing wind/solar and 5) good subsidy support. However, the figures presented in this article (Lazard) are over-optimistic, using low-end capital costs and high-end capacity factors. As shown below, real world numbers for wind and solar costs result in LCOE numbers which are 32% and 64% higher than Lazard’s estimates respectively. 

Let’s compare Lazard’s 2012 analysis to real world data for wind and solar.

Wind:

Lazard: capital costs: $1500-2000, capacity factor: 30-48%

Real world: capital costs: $2000/kW average ($1400-4400/kW), capacity factor: 33% average (18-53%)

LCOE using 20 year lifetime, 6% discount rate and operating costs of $30/kW/yr:

Lazard $1750/kW and 39%: $53.4/MWh

Real world $2000/kW and 33%: $70.7/MWh (32% higher)

Solar:

Lazard: capital costs: $2000-2750/kW, capacity factor: 20-27%

Real world: capital costs: $4000/kW average ($2600-8000/kW), capacity factor 23% average (15-30%)

For PV it is important to distinguish between AC and DC capacity factors and costs, something which Lazard seems to have missed. 

LCOE using 30 year lifetime, 6% discount rate and operating costs of $30/kW/yr:

Lazard $2400/kW and 24%: $97.2/MWh

Real world $4000/kW and 23%: $159.1/MWh (64% higher)

Bruce McFarling's picture
Bruce McFarling on February 14, 2015

Renewable energy can’t compete with conventional energy as to the net cost of displacing C02 because it is intermittent.”

This is where knowledge of statistical frequency analysis comes in handy, since it is straightforward that as the renewable energy portfolio is expanded in terms of (1) the distribution of wind turbines across a windpower supply region, (2) transmission connecting grids with independent windpower supply, (3) the distribution of solar power across a solar supply region, (4) transmission connecting grids across different time zones and with independence of reductions of solar incidence and (5) combining wind and solar on the same grid, the share of the RE energy portfolio that is intermittent declines, and the share that is not intermittent increases.

So the premise “intermittence rules out deriving a large share of our energy supply from the harvest of variable RE” requires the tacit appendix, “so long as we try to harvest that supply from a too narrow area and too limited a range of different variable RE sources of energy, and refuse to include any substantial share of dispatchable RE in the portfolio”.

And with the appendix, instead of being “why RE cannot be used”, it becomes a description of how not to use RE.

Bruce McFarling's picture
Bruce McFarling on February 14, 2015

The experience in France is that relying on a very heavy share of nuclear also requires a substantial back-up capacity of dispatchable fueled energy sources … and of course France is not an island, and the heavy French use of nuclear is also supported by cross-haul transmission with the rest of Europe.

Willem Post's picture
Willem Post on February 14, 2015

Deleted duplicate

Willem Post's picture
Willem Post on February 14, 2015

Schalk,

The numbers in the article are grossly underestimated, as you show, and would be even more so, if the below factors were accounted for.

The Northeast has an overall capacity factor of 0.25. Ridgeline wind costs about $2,600,000/MW + grid connection costs.

– The 63 MW ridgeline wind system in Vermont took $20 million to properly connect to the grid, plus about 1.5 years of lesser production to minimize grid disturbances.

– The 60 MW Seneca project was cancelled because $86 million of grid upgrades were required.

Another cost is adjusting the generator population to have the mix of capacities to adequately balance the variable energy. That new mix of capacities likely would not be as optimum as without variable RE.

Another cost is the lesser efficiency of the generators that are doing and/or are impacted by the balancing.

Another cost is having nearly ALL conventional generators available, in good working order, staffed and fueled, when there is no or minimal wind and sun, which can occur during any time of the year.

http://theenergycollective.com/willem-post/310631/more-realistic-cost-wi...

http://theenergycollective.com/willem-post/332911/high-renewable-energy-...

http://theenergycollective.com/willem-post/338781/high-renewable-energy-...

Nathan Wilson's picture
Nathan Wilson on February 15, 2015

Yes, zero-carbon grids built with either nuclear or variable renewables would both involve use of long distance transmission, dispatchable backup power, and perhaps energy storage; the difference is that the nuclear-rich portfolio would need about an order of magnitude less.  

The reason is that nuclear plants fail independently (e.g. a grid with 100  nuclear plants would never have more than a few percent of them fail during peak demand times), whereas all wind farms in a large region can fail simultaneously due to weather systems, and of course all solar plants fail simultaneously every night.

Willem Post's picture
Willem Post on February 15, 2015

Bruce,

Can you support that statement with some back up?

Charlie Hewitt's picture
Charlie Hewitt on February 16, 2015

Texas has seen incredible growth in its wind generation installed capacity.  According to ERCOT, wind ICAP was 1173 MW in 2003 and grew to 11,065 MW by the end of 2013. 

Completion of the $7 billion Competitive Renewable Energy Zone will eventually bring 18,500 MW of wind capapcity from the West Texas wind farms to the state’s population centers.  To examine a levelized cost of energy there must be some accounting for the transmission construction costs necessary to realize the benefit of the generation. 

Also, it is interesting to note that as West Texas winds are noted for blowing contrary to when demand is peaking, the development of projects in coastal areas has increased dramatically.  Coastal winds are more reliable and are typically strongest when system demand peaks.  In fact, based on multi-year averages, ERCOT has assigned a summer peak average capacity for coastal wind of 56% compared to only 12% for non-coastal wind.

As others have already stated more eloquently, if you are going to compare costs, you need be certain 1) you have included all assignable costs, and 2) the capacity is actually available when it is needed.  If you cannot use two sources of generation interchangably, any cost comparisons are somewhat lacking in relevance.

Wayne Lusvardi's picture
Wayne Lusvardi on February 16, 2015

Bruce

If I understand your comment correctly, you are saying that the statement “RE is intermittent” must be qualified because if there were a gigantic RE system across time zones in the U.S. then intermittency would be reduced.  

But if that was the case, then so would costly redundancy.  Marc Jacobson of Stanford University’s Precourt Energy Institute has concocted a plan whereby RE would supply 100% of all energy needs. But it would require an energy system, by my guess, at least four times the existing system in order to provide any reliability.  

I worked for a large regional water agency in California for 20 years. I learned that water and energy systems have to have redundancy built in especially in the event of a catastrophic loss of part of the grid or many power plants simultaneously.  What I am talking about is the redundancy needed to make RE reliable 24/7/365 as does, say, nuclear power. 

 

Wayne Lusvardi's picture
Wayne Lusvardi on February 16, 2015

Bruce

If I understand your comment correctly, you are saying that the statement “RE is intermittent” must be qualified because if there were a gigantic RE system across time zones in the U.S. then intermittency would be reduced.  

But if that was the case, then so would costly redundancy.  Marc Jacobson of Stanford University’s Precourt Energy Institute has concocted a plan whereby RE would supply 100% of all energy needs. But it would require an energy system, by my guess, at least four times the existing system in order to provide any reliability.  

I worked for a large regional water agency in California for 20 years. I learned that water and energy systems have to have redundancy built in especially in the event of a catastrophic loss of part of the grid or many power plants simultaneously.  What I am talking about is the redundancy needed to make RE reliable 24/7/365 as does, say, nuclear power. 

 

Bruce McFarling's picture
Bruce McFarling on February 16, 2015

I am saying that the statement “RE is intermittent” has to be qualified by looking at real world output duration curves and coveriances among different sources and between different portfolios and demand.

Fertig (2012 ~ this is a poster, as the article is now behind a paywall) find that interconnection of the BPA, ERCOT, MISO and CAISO would double the amount of firm power (available 92% of the time) compared to the the best case of the four (MISO). On the order of 15% of nameplate capacity would be firm power, so with a capacity factor in the range of 30%-35%, over 40% of power generated from wind by that system would be firm power.

As can be seen in the output duration curve from Fertig (2012), BPA wind on its own is essentially all intermittent, but since it is effectively independent in the covariance of supply with MISO, interconnecting MISO and BPA would still reduce the intermittence of MISO windpower.

That is windpower alone, but HVDC transmission is agnostic to the source of the electrons, and that can just as easily be used to shave the top off of peak solar generating periods in Central Texas two hours before Southern California solar peaks and from Southern California two hours after Central Texas peaks.

As Fertig (2012) notes, at present (subsidized) costs for NG electricity generation, that firming is financially cheaper to provide with NG. However, on a full economic cost analysis including the full cost of the NG emissions the transmission would be economically competitive.

 

Bruce McFarling's picture
Bruce McFarling on February 16, 2015

The redundancy needed to make firm RE as reliable as nuclear power is similar to the redundancy needed to make nuclear power as reliable as nuclear power.

Since nuclear power is not, in fact, 24/7/365, in either case, substantial dispatchable capacity is required to compensate for either the predictable (predicted changed in windspeed, planned reactor downtimes for fuel changes) or unpredictable component of supply variability. The nuclear requires a smaller total energy supply and fast dispatch for a larger share of dispatchable back up power, since large unpredicted supply outages are not common, but tend to be rapid when they occur, while windpower / solarpower / run-river-hydro / etc. requires larger total energy supply but a smaller share requires fast dispatch, since on the order of half of variability of RE supply can be predicted a day or more in advance.

 

Get Published - Build a Following

The Energy Central Power Industry Network is based on one core idea - power industry professionals helping each other and advancing the industry by sharing and learning from each other.

If you have an experience or insight to share or have learned something from a conference or seminar, your peers and colleagues on Energy Central want to hear about it. It's also easy to share a link to an article you've liked or an industry resource that you think would be helpful.

                 Learn more about posting on Energy Central »