Cutting Back :: Plans to Lessen Dependence on Fossil Fuels

[barbos]

Maryland does not have to generate all electricity they use. You need only part of the Mojave desert to generate all electricity US use.

You mean with solar panels?

yes

 

[JonA]

Just for fun, I ran some numbers on this for the state of Maryland:

• 56,400,000,000 miles were traveled in the state in 2014 (source).
• High estimates for EV efficiency give 6 miles/kWh; lower estimates give 2 miles/kWh (source - PDF).
• Based on that, a conversion of Maryland's vehicles to electricity would require generation of an additional 9,400,000 mWh/yr - 28,200,000 mWh/yr of electricity.*

__________
* That there are already some EVs on the road was considered but not included in the calculation though it would reduce the 'additional' since some of today's electricity is already being used to power EVs. However, the number of current vehicle miles being traveled with EVs is likely less than 1% based on national averages, making the effect possibly negligible.

First. Good posts JonA.

Just thought I'd bring out my maintainability and sustainability wrench and see what I could do.

Governor's Guide to Moderinizing Electric Power Grid (Maryland) http://www.nga.org:8080/files/live/...overnorsGuideModernizingElectricPowerGrid.pdf

illustrates what I see as a hole you left in your analysis. Age of average grid system is over 25 years, generators over 30 years, etc, and most designed in '50s found in summary are inline with my experience here in the NW. From the mid forties to the mid fifties my dad was heavily involved in constructing the Bonneville grid. Not much has happened since. The systems are too old already and need remediation just to sustain their near term demands. So add replacing to the bill for increasing.

Thank you for the source.

To add to that, all that replacing is likely to be done with men in trucks and machines that burn lots of fossil fuels as they move around cables forged in factories relying on lots of fossil fuels. And then they're building towers and running those cables, perhaps through otherwise pristine environments (see this recent BBC article about wilderness destruction). So a comprehensive analysis of the overall benefit (just to the environment) of switching to EVs will have to include those costs of the project.

 

[JonA]

You mean with solar panels?

yes

Oh.

Then you are absolutely wrong. The U.S. used 4,224,000,000 MWh of electricity in 2011 (source). The large, much-admired Topaz plant (which I used in my calculations for the EvC post) generates 1,100,000 MWh per year in an area of 9.5 sq miles.

It would take 3,840 Topaz plants covering an area of 36,000 sq miles to generate the U.S.'s electricity for a year.

The Mojave Desert is 25,000 sq miles (source).

 

[barbos]

yes

Oh.

Then you are absolutely wrong. The U.S. used 4,224,000,000 MWh of electricity in 2011 (source). The large, much-admired Topaz plant (which I used in my calculations for the EvC post) generates 1,100,000 MWh per year in an area of 9.5 sq miles.

It would take 3,840 Topaz plants covering an area of 36,000 sq miles to generate the U.S.'s electricity for a year.

The Mojave Desert is 25,000 sq miles (source).

I really don't care for your calculation using some Topaz plant.
My calculations based on 10% efficient PV gave me 10,000 km^2. that's 15% of that desert.

 

[bigfield]

... a 30% increase in generation is achievable.

Yes. But how are we planning to generate it?

If it's fossil fuels, the answer is self-evident. But if we are using solar panels, things are not so obvious. Most drivers will be charging their cars at night, which means you have to capture the electricity and store it for charging later. In fact the need for storage is pretty much built in to any plan for wide-scale renewable energy generation. And storage is difficult and expensive. That is why today, September 9, 2016, we know of no advanced societies (= U.S., Canada, Europe, etc.) that rely on renewables such as solar and wind.

Yes, it is difficult and expensive, but certainly achievable.

That means that if you switch to electric vehicles this instant, you're almost certainly going to be meeting a large portion of the demand for their electricity with fossil fuels. Now that may still be alright. I might go looking for the numbers later, but it's possible that using fossil fuels to generate electricity that powers a car is more efficient (and thus less polluting, etc.) than using the fossil fuels directly in the car - even when comparing fossil fuels such as dirty coal with cleaner petroleum. And we can explore the same issue for the whole lifecycle of the fuel and generation method in question to really compare the overall benefits and costs of the different scenarios.

Let's assume that grid electricity from fossils is dirtier than petrol/gasoline: that problem would only last until the percentage of electricity from fossil fuels is reduced to invert that relationship.

Power supplied to the grid would become cleaner simply by supplying the additional demand from EVs with new clean generators. If that isn't enough of a reduction then the gradual retirement of fossil-fuel plants will do the job.

 

[JonA]

Oh.

Then you are absolutely wrong. The U.S. used 4,224,000,000 MWh of electricity in 2011 (source). The large, much-admired Topaz plant (which I used in my calculations for the EvC post) generates 1,100,000 MWh per year in an area of 9.5 sq miles.

It would take 3,840 Topaz plants covering an area of 36,000 sq miles to generate the U.S.'s electricity for a year.

The Mojave Desert is 25,000 sq miles (source).

I really don't care for your calculation using some Topaz plant.
My calculations based on 10% efficient PV gave me 10,000 km^2. that's 15% of that desert.

How exactly are you making your calculations?

I cannot get anywhere close to your values even making best-case assumptions. For example, the most efficient solar cells (in terms of land use) have a capacity-land ratio of 175W/m² (source). If we're operating at 10% efficiency to get the power required for the U.S., we'd need an installation with a 4,821,917 MW capacity. Even using the highest efficiency numbers from the Wiki link above, that's still over 25,000km² (~10,500 sq miles). In fact, even using the efficiency figures (in terms of output) from Topaz of ~25%, I still get 11,000km² (~4,250 sq miles).

And none of that even includes other components of generating useful electricity from solar panels that are required in a 'plant' installation, which is the reason I used an operating plant like Topaz as a starting point in the first place instead of base measurements of individual solar panels.

So what are you using to get your numbers?

As far as I can see, there's simply no realistic way to make your 10,000km² estimate work.

 

[JonA]

Yes. But how are we planning to generate it?

If it's fossil fuels, the answer is self-evident. But if we are using solar panels, things are not so obvious. Most drivers will be charging their cars at night, which means you have to capture the electricity and store it for charging later. In fact the need for storage is pretty much built in to any plan for wide-scale renewable energy generation. And storage is difficult and expensive. That is why today, September 9, 2016, we know of no advanced societies (= U.S., Canada, Europe, etc.) that rely on renewables such as solar and wind.

Yes, it is difficult and expensive, but certainly achievable.

That means that if you switch to electric vehicles this instant, you're almost certainly going to be meeting a large portion of the demand for their electricity with fossil fuels. Now that may still be alright. I might go looking for the numbers later, but it's possible that using fossil fuels to generate electricity that powers a car is more efficient (and thus less polluting, etc.) than using the fossil fuels directly in the car - even when comparing fossil fuels such as dirty coal with cleaner petroleum. And we can explore the same issue for the whole lifecycle of the fuel and generation method in question to really compare the overall benefits and costs of the different scenarios.

Let's assume that grid electricity from fossils is dirtier than petrol/gasoline: that problem would only last until the percentage of electricity from fossil fuels is reduced to invert that relationship.

Power supplied to the grid would become cleaner simply by supplying the additional demand from EVs with new clean generators. If that isn't enough of a reduction then the gradual retirement of fossil-fuel plants will do the job.

Thing is, I agree with you on a gut-feeling emotional level. But realistically I just don't know (and I think very few people do). And that makes me hesitant to start championing a complete change up that could potentially make things much worse than they already are.

That's why before throwing my support behind these alternatives, I want to get the evidence that shows they are really the right way to go.

 

[barbos]

I really don't care for your calculation using some Topaz plant.
My calculations based on 10% efficient PV gave me 10,000 km^2. that's 15% of that desert.

How exactly are you making your calculations?

I cannot get anywhere close to your values even making best-case assumptions. For example, the most efficient solar cells (in terms of land use) have a capacity-land ratio of 175W/m² (source). If we're operating at 10% efficiency to get the power required for the U.S., we'd need an installation with a 4,821,917 MW capacity. Even using the highest efficiency numbers from the Wiki link above, that's still over 25,000km² (~10,500 sq miles). In fact, even using the efficiency figures (in terms of output) from Topaz of ~25%, I still get 11,000km² (~4,250 sq miles).

And none of that even includes other components of generating useful electricity from solar panels that are required in a 'plant' installation, which is the reason I used an operating plant like Topaz as a starting point in the first place instead of base measurements of individual solar panels.

So what are you using to get your numbers?

As far as I can see, there's simply no realistic way to make your 10,000km² estimate work.

I did the estimate long time ago and now I think i used 20% efficiency.
Assuming 300mil people, 2 kW per person average , 100% sunny (1 kw/m^2) desert close to equator. Then area will be:
2/0.2*M_PI=10*Pi=31.4159265358979 m^2 per person,
for 300mil people it's about 9425 km^2

 

[JonA]

How exactly are you making your calculations?

I cannot get anywhere close to your values even making best-case assumptions. For example, the most efficient solar cells (in terms of land use) have a capacity-land ratio of 175W/m² (source). If we're operating at 10% efficiency to get the power required for the U.S., we'd need an installation with a 4,821,917 MW capacity. Even using the highest efficiency numbers from the Wiki link above, that's still over 25,000km² (~10,500 sq miles). In fact, even using the efficiency figures (in terms of output) from Topaz of ~25%, I still get 11,000km² (~4,250 sq miles).

And none of that even includes other components of generating useful electricity from solar panels that are required in a 'plant' installation, which is the reason I used an operating plant like Topaz as a starting point in the first place instead of base measurements of individual solar panels.

So what are you using to get your numbers?

As far as I can see, there's simply no realistic way to make your 10,000km² estimate work.

I did the estimate long time ago and now I think i used 20% efficiency.
Assuming 300mil people, 2 kW per person average , 100% sunny (1 kw/m^2) desert close to equator. Then area will be:
2/0.2*M_PI=10*Pi=31.4159265358979 m^2 per person,
for 300mil people it's about 9425 km^2

Your numbers are off - by a lot. The US used 4,224,000,000 MWh of electricity in 2011 - that is a fact. At 20% generating efficiency, you'd need an installed capacity of 2,400,000 MW. Assuming your 300 million population, that's an installed capacity of 8kW/person.

And with only 2kW capacity per person, at 20% efficiency, you're talking about an output of 3,500kWh/person/year. Not even a country like Libya has a consumption rate that low.

Also, why are you multiplying anything by pi? A solar installation has a given size already calculated in terms of area. To properly use your inputs to find the land area requirements, we have to do the following math - unfortunately containing no kind of pi(e):

2kW/person * 300,000,000persons = 600,000,000kW capacity
600,000,000kW / 1kW/m² = 600,000,000m²
600,000,000m² = 600km²​

That number, of course, is way off because 1kW/m² is outrageously more than what we can currently get from solar cell technology. If we were using a more reasonable figure of 200W/m², the estimate looks something more like:

2kW/person * 300,000,000persons = 600,000,000kW capacity
600,000,000kW / 0.2kW/m² = 3,000,000,000m²
3,000,000,000m² = 3,000km²​

To get it even more realistic, we use a capacity requirement more in line with how much electricity the U.S. actually uses, giving us:

8kW/person * 300,000,000persons = 2,400,000,000kW capacity
2,400,000,000kW / 0.2kW/m² = 12,000,000,000m²
12,000,000,000m² = 12,000km²​

In fact once we've replaced all of our inputs with ones properly grounded in reality, we see that the estimate of 10km² is simply not accurate. Your number is kind of close, though; which is a miracle considering the math you used to get there.

ETA: To be fair, if you did these calculations long enough ago, your estimates of electricity use may be more reasonable. The U.S. was probably using 3,500kWh/person/year somewhere between the 1940s-1960s. That's definitely not the case today though.

 

[barbos]

I did the estimate long time ago and now I think i used 20% efficiency.
Assuming 300mil people, 2 kW per person average , 100% sunny (1 kw/m^2) desert close to equator. Then area will be:
2/0.2*M_PI=10*Pi=31.4159265358979 m^2 per person,
for 300mil people it's about 9425 km^2

Your numbers are off - by a lot. The US used 4,224,000,000 MWh of electricity in 2011 - that is a fact. At 20% generating efficiency, you'd need an installed capacity of 2,400,000 MW. Assuming your 300 million population, that's an installed capacity of 8kW/person.

your number is actually lower than one I used
4224000000000/300e6/365/24. = 1.60730593607306 kW per person.
My calculation assume 2 kW per person.

In fact once we've replaced all of our inputs with ones properly grounded in reality, we see that the estimate of 10km2

My estimate is 10,000 km^2

Also, why are you multiplying anything by pi?

That's how geometry works.

 

[JonA]

Your numbers are off - by a lot. The US used 4,224,000,000 MWh of electricity in 2011 - that is a fact. At 20% generating efficiency, you'd need an installed capacity of 2,400,000 MW. Assuming your 300 million population, that's an installed capacity of 8kW/person.

your number is actually lower than one I used
4224000000000/300e6/365/24. = 1.60730593607306 kW per person.
My calculation assume 2 kW per person.

I think you missed the point of my critique, which was that given the amount of electricity used, at 20% efficiency, we need an installed capacity of 8kW/person.

If we operated at 100% efficiency 24/7, yes, an installed capacity of 2kW/person would do the trick.

But we don't, and you even said so. A capacity of 2kW/person at 20% efficiency gives an output of 3,500kWh/person/year. To get the real value of 14,000kWh/person/year requires a capacity of 8kWh/person/year at 20% efficiency.

There's no way around that.

In fact once we've replaced all of our inputs with ones properly grounded in reality, we see that the estimate of 10km2

My estimate is 10,000 km^2

Correct. I mistyped.

Also, why are you multiplying anything by pi?

That's how geometry works.

 

[barbos]

your number is actually lower than one I used
4224000000000/300e6/365/24. = 1.60730593607306 kW per person.
My calculation assume 2 kW per person.

I think you missed the point of my critique, which was that given the amount of electricity used, at 20% efficiency, we need an installed capacity of 8kW/person.

If we operated at 100% efficiency 24/7, yes, an installed capacity of 2kW/person would do the trick.

But we don't, and you even said so. A capacity of 2kW/person at 20% efficiency gives an output of 3,500kWh/person/year. To get the real value of 14,000kWh/person/year requires a capacity of 8kWh/person/year at 20% efficiency.

Where does this another 20% come from? I already accounted for 20% efficiency of PV panel.

There's no way around that.

Since I don't see that I don't need to around it

In fact once we've replaced all of our inputs with ones properly grounded in reality, we see that the estimate of 10km2

My estimate is 10,000 km^2

Correct. I mistyped.

Also, why are you multiplying anything by pi?

That's how geometry works.

Well, Sun is not always at 90 degrees to surface of the earth, and half of the time it's night. so if you do the math you will find out that you need your solar panel be Pi times bigger than the case where it is 90 degrees 24 hours a day.

 

[JonA]

I think you missed the point of my critique, which was that given the amount of electricity used, at 20% efficiency, we need an installed capacity of 8kW/person.

If we operated at 100% efficiency 24/7, yes, an installed capacity of 2kW/person would do the trick.

But we don't, and you even said so. A capacity of 2kW/person at 20% efficiency gives an output of 3,500kWh/person/year. To get the real value of 14,000kWh/person/year requires a capacity of 8kWh/person/year at 20% efficiency.

Where does this another 20% come from? I already accounted for 20% efficiency of PV panel.

There's no way around that.

Since I don't see that I don't need to around it

In fact once we've replaced all of our inputs with ones properly grounded in reality, we see that the estimate of 10km2

My estimate is 10,000 km^2

Correct. I mistyped.

Also, why are you multiplying anything by pi?

That's how geometry works.

Well, Sun is not always at 90 degrees to surface of the earth, and half of the time it's night. so if you do the math you will find out that you need your solar panel be Pi times bigger than the case where it is 90 degrees 24 hours a day.

I believe I see what's going on.

Solar irradiance is 1kW/m². That's where you've been starting from. The best solar panels can only turn about 180W of that 1kW into electricity. That's the first measure of efficiency and where I've been starting from. The second measure of efficiency is how often the panels can put out at that peak. You're guessing at that number using a formula for sun angle, etc., whereas I'm using efficiencies gathered from actual usage of these systems. This, I believe, is accounting for our difference in 9,400km² and 11,000km². I can concede that for the panels alone a figure of 10,000km² isn't outrageous.

So now that we know the panels can fit - and how much space they require - , we can look at the other bits.

As it sits, the US has an installed capacity of a little over 1,000GW (source). If those sources ran at capacity for a year, they'd generate 8,760,000,000MWh of electricity, meaning our current grid operates on a roughly 2:1 ratio of max production to actual consumption. There are several reasons for this, one of which is reliability; the other is the need to satisfy peaks in demand. On the latter, for example, peaks of 700,000MWh or more are not uncommon (source).

The system you've described would max out, however, at around 400,000MW, which means we'd need either more panels or a regulatory storage system. Since we need a storage system anyway given that no panels produce at night, that's probably the best place to look next.

Do you know how much storage we'll need?

 

[Angra Mainyu]

... a 30% increase in generation is achievable.

Yes. But how are we planning to generate it?

If it's fossil fuels, the answer is self-evident. But if we are using solar panels, things are not so obvious. Most drivers will be charging their cars at night, which means you have to capture the electricity and store it for charging later. In fact the need for storage is pretty much built in to any plan for wide-scale renewable energy generation. And storage is difficult and expensive. That is why today, September 9, 2016, we know of no advanced societies (= U.S., Canada, Europe, etc.) that rely on renewables such as solar and wind.

That means that if you switch to electric vehicles this instant, you're almost certainly going to be meeting a large portion of the demand for their electricity with fossil fuels. Now that may still be alright. I might go looking for the numbers later, but it's possible that using fossil fuels to generate electricity that powers a car is more efficient (and thus less polluting, etc.) than using the fossil fuels directly in the car - even when comparing fossil fuels such as dirty coal with cleaner petroleum. And we can explore the same issue for the whole lifecycle of the fuel and generation method in question to really compare the overall benefits and costs of the different scenarios.

How about nuclear (+ solar, wind, hydroelectric, etc.)?

It would be more expensive than fossil fuels, but much less of a problem for the environment to just make enough advanced nuclear reactors to increase generation beyond what solar, etc., can reliably provide with present-day tech (at a cost that might also be high but not unachievable). As electric cars, buses, trucks, etc., become more common, they can be powered by electricity generated by nuclear reactors.

As time goes by and new solar technologies, storage technologies, etc., are developed, nuclear fission (at least on Earth) can be gradually replaced by solar and perhaps some other sources. It seems a lot better than present-day generation, all other things equal.
Of course, not all other things are equal, e.g., money isn't, and I've not done the math, so I don't know if it's overall a good idea; I'm just suggesting a potential alternative.
By the way, China seems decided to make a lot of new reactors, and electric cars are coming, so it seems to me that sooner or later, at least in parts of the world this will happen (yes, China is still making coal power stations, but increasing the amount of smog likely increases political instability; at some point, the government will probably stop that out of self-interest).

 

[barbos]

Solar irradiance is 1kW/m2. That's where you've been starting from. The best solar panels can only turn about 180W of that 1kW into electricity

That would make solar panels 18% efficient. I am using round 20% number. and best panels are I think around 40% efficient, they are 100 times more expensive but they exist.
Now if you accept my assumptions then you would get 10,000 km^2. Of course you will need a storage, how much will depends on how much you charge people for a night electricity which in turn will depend on the cost of that storage.