ALL Are Here to Stay

Today, in the energy discourse there are constant skirmishes between the proponents of different types of energy.

Depending on the respective camp, people want to eliminate (pick one) fossil fuels, nuclear, wind, solar, what have you.

Well, we have news for all the groups: NONE of the current energy sources is going away, at least not this century. Whether we like it or not, on a global scale, we'll have to learn to live with them.

Now, after saying the above, it doesn't mean we should write a blank check to any particular technology without FIRST doing our homework.

As an example, let's analyse solar photo-voltaic (PV).

What is the highest PV penetration that makes economic / environmental sense in a national grid?*

Since PV is intermittent, it probably makes no sense to go much above low single digits. Why? Let's do our homework with an example:

Say a country uses, on average, 40 GW of electricity. At peak hour, they consume 40% more than the average, in other words, 56 GW.

Thus, the maximum output of the solar panels at any particular moment should not exceed 56 GW (unless we want to embark in expensive / environmentally challenging massive storage which today is not ready for prime time).

Consequently, the PV installed capacity in this country should be capped at 56 GW.

If the solar annual capacity factor in this country is 15%, then the average annual production of the PV installation will be: 56 GW x 0.15 = 8.4 GW.

The country itself consumes 40 GW average, so the PV component would be: 8.4 / 40 = 21%.**

The above means that at peak solar production ALL other generating capacity would need to be idled / shut down. At night (and to a lesser extent during cloudy days), the other types of generators would have to supply the electricity requirements.***

Non intermittent power sources such as fossil fuels and nuclear are not constrained by the above mentioned "cap."

Hydro is somewhere in the middle since its intermittency is not daily but seasonal or from year to year.

Conclussion: intermittent energy has a "natural" cap that would make no economic / environmental sense to exceed.

Feel free to add to the conversation on Twitter: @luisbaram


Notes:
* Sure, one country could "dump" excess power into another, but if that second country uses the same type of technology to produce its electricity, they would have surpluses at the same time.

** This is really an optimistic number since, for example, in Europe more electricity is required in winter when solar produces the least energy. I propose the "rule of thumb" for solar should be to cap it at the annual capacity factor. e.g. if the annual capacity factor is 15%, then at the most 15% of the annual electricity should be solar. However, even this number might be too high.

*** The costs per GWh of the modulated / idled / shut down power plants are higher than if they could produce continually at their capacity. These costs ultimately affect the overall prices of the electricity in the grid.

Solar Photo-voltaic: Let's Get Real

This exercise is a simplification in which we try to explain why solar photo-voltaic (PV) has difficulty in living up to its hype.

Let's consider an isolated country that consumes on average 40 GW of electricity.

To simplify things, let's consider they consume this amount of power 24 hours a day. In real life, consumption has peaks and valleys.

They plan to install enough solar PV to supply 100% of the electrical energy of the country at peak solar production.

Again, to simplify, let's consider a perfectly cloudless day during the spring or fall equinox. The output would look like this:

So, from midnight to 6 AM, another energy source would supply 100% of the electricity.

Then again, from 6 PM until midnight, another energy source would supply 100% of the electricity.

Then, from 6 AM until 6 PM solar would provide continually variable production and will reach 100% of this country's energy needs at local noon. In other words, at local noon, solar PV would be producing 40 GW of power. (8)

What would happen if this country decided to go above 100% solar at peak production (as an isolated country, they couldn't "dump" the excess production into another country. Also, we are not considering storage that could be an article in itself).

Then, they would have to curtail (disconnect) solar capacity at peak production hours. This is how the graph would look (100% peak vs. 125% peak comparison):

As we may see, there is not too much sense in taking the PV capacity above the peak requirements.

Now, how would the production of the "other" sources (usually fossil fuels) look to compensate for the variable nature of PV. Here we can see it:

In other words, from midnight to 6 am, and from 6 pm until midnight, the other source would supply 100% of the electricity. Then from 6 am to 6 pm it would have to continually adjust its output to compensate for the PV production. 

If the Earth were a perfectly cloudless planet, this sort of arrangement would allow PV to provide close to one third the energy requirements of a country. What would be the carbon intensity of such electricity? Here we calculate it:

According to the table referenced below, solar PV has a carbon intensity of 46 grams per kWh, and let's say the rest of the electricity is produced by natural gas (469 grams per kWh), thus the combined carbon intensity would be:

     46 x 0.33 + 469 x 0.67 = 329 grams per kWh.

However, in real life the Earth is not cloudless and thus the actual annual capacity factor of solar PV is closer to 15%. If we re-calculate with this more realistic number, we get:

     46 x 0.15 + 469 x 0.85 = 406 grams per kWh.

If a component of coal is used in the "other" energy then the emissions would rise even higher. 

Again, this article is a simplification, but the point is to explain in simple terms why solar PV is not living up to its hype.

Thank you.

Notes:

1. In real life, clouds reduce the output of the solar panels.

2. Seasonality also greatly impacts power generation: winter days are shorter and possibly cloudier.

3. The "other" power plants need to be idled, modulated, shut down, restarted and this causes inefficiencies in the system and additional emissions.

4. From a purely operational point of view, "nothing would happen" if all solar capacity were disconnected.

5. Yes, a solar + fossil fuels system produces less emissions than a purely fossil fuel one, but at the cost of duplicated investment.

6. Yes, excess solar energy could be "dumped" into another country, but if that country also installed significant solar capacity, this wouldn't be an option anymore. 

7. The other option is storage but currently this (expensive) technology hasn't been widely deployed. Also, storage would add to the emissions per kWh (once life-cycle emissions are taken into consideration).

8. "Local noon" doesn't happen at the same time in all the country, so the curve would be a little bit flattened. 

References:

http://gnwr1.blogspot.mx/2013/01/clean-energy.html 

Going Nuclear

This is a simplified exercise to visualize what would be required to convert 50% of the world's electricity to nuclear, the premier low carbon source.

According to the EIA (Energy Information Administration) latest report, the global generation of electricity in 2010 was 20.2 trillion kWh, and they project that by 2040 total global generation will reach 39.0 trillion kWh.

So, if we decide to generate 50% of our electrical energy with nuclear (at, say, an average 80% capacity factor), we would require this number of 1 GWe reactors:

     50% of 39.0 trillion kWh is 19.5.

Let's convert trillion kWh to GWh by multiplying it by one million:

     We now have 19,500,000 GWh.

A 1 GWe nuclear plant at 80% capacity factor (CF) produces, on an annual basis:

     1GWe x 0.80CF x 24 hrs x 365 days = 7,008 GWh.

Thus, we would need this number of reactors by 2040:

     19,500,000 / 7,008 = 2,783.

Simplifying thing a little, let's consider that half of the current nuclear plants will still be in operation by 2040. According to the EIA, nuclear supplied 2.6 trillion kWh in 2010. This would correspond to the equivalent of 371 1 GWe reactors at 80% CF. If half of these are in operation by 2040, we can subtract 186 from the number calculated above, thus we get:

     2,783 - 186 = 2,597 new nuclear reactors.

If we have 30 years to build them, it would require the commissioning of:

     2,597 / 30 = 87 nuclear reactors EVERY year for 30 CONSECUTIVE years.

And again, let's remember that the above effort would only yield 50% of our global electricity and around 25% of our total energy consumption by 2040.

At the end of the day we have to differentiate what is possible, from what is probable.

The probability of this nuclear build up taking place by 2040 is, in my humble opinion, less than zero.

Feel free to add to the conversation.

Thank you.

EIA electricity projections from their latest report:

Notes:
1. By 2040, only ~50% of our energy consumption will be electricity.
2. There is no such thing as a global grid, so the real life implementation would be more complicated than pictured above.
3. Sure, many questions need to be answered, starting with determining if we even have the manufacturing capacity required to make such a ramp up.
4. Nuclear has a serious advantage over RE (sun and wind): it is baseload, reliable power.
5. Of the renewable power in 2040, fully 65% is estimated to be hydro.
6. Yes, there are reactors bigger and smaller than 1 GWe, we are considering all of them at 1 GWe to simplify.