Go Back  The Hull Truth - Boating and Fishing Forum > BOATING FORUMS > Dockside Chat
Reload this Page >

How many PV solar panels would it take...?

Notices
Like Tree26Likes

How many PV solar panels would it take...?

Old 04-26-2019, 04:50 AM
  #1  
Senior Member
Thread Starter
 
Join Date: Jun 2014
Location: Virginia
Posts: 1,761
Default How many PV solar panels would it take...?

How many PV Solar panels would it take to power the US?

The great and all knowing pot smoker, Elon Musk claims an area in Arizona of approximately 100 sq/mi with a battery farm of 1mi. by 1mi. (does not specify technology or height of the battery farm) would be capable of powering the US 24/7/365 based on current power consumption. I'm calling BS on it. Anyone else see some issues with this like peak demand, inclement weather events, transmission losses, overheating inefficiencies, seasonal demand changes, etc. I think he is using a lot of averages based on base load calculations. Might work on paper but not in the real world.



Interestingly, I haven't seen him address how everyone converting to EV's will impact his proposed solar grid size especially considering most of the charging would likely be when the sun isn't shining.

Copying the math of someone else who gave this some thought it would take an additional 500 sq. miles to provide the power necessary to power the entire US vehicle fleet. These two combined would be removing just the vehicles and utility electric generators from the equation with a grid of 600 sq/mi. or roughly 3X the surface area of Lake Mead. Apparently he thinks the energy potential of this solar farm is significantly more potent than the potential energy storage of hydroelectric because Lake Mead has trouble meeting the power needs of just 1.3M people and we have 327M in the US if you don't count the undocumented illegals. It speaks nothing to other consumptions of fossil fuels such as shipping, home heating, cooking, industrial and agricultural use of NG, LP, and diesel, recreational and lawn care products, etc.

Curious as to others thoughts on what a working solar model would look like. How do you power places like Hawaii, Alaska, or even the northeast population centers, year around using solar power from Arizona?__________________
"All any man needs, is five things-good health, a good job, a good wife, a good religion and at least one goldarn good hobby." - Gar Wood
1blueheron is offline  
Old 04-26-2019, 05:16 AM
  #2  
Senior Member
 
Join Date: Oct 2015
Posts: 995
Default

Some very rough back of a soggy napkin calculation , assuming a 250 watt panel will give you a yearly average of 1KWh per day, and takes 2 square meters of space, current US electrical demand could be covered by ~32 square miles of panels.

You would also need about 80 gigawatt hours of battery....

to makeit weather resistant, triple the amount of panels, and increase the batteries by at least 5 times....

replace all your cars with ev’s, increase the above by an order of magnitude as a starting point...
mickt243 is offline  
Old 04-26-2019, 05:39 AM
  #3  
Senior Member
Thread Starter
 
Join Date: Jun 2014
Location: Virginia
Posts: 1,761
Default

Originally Posted by mickt243 View Post
Some very rough back of a soggy napkin calculation , assuming a 250 watt panel will give you a yearly average of 1KWh per day, and takes 2 square meters of space, could be covered by ~32 square miles of panels.

You would also need about 80 gigawatt hours of battery....
What figure are you using for "current US electrical demand"? Your system won't meet demands in the winter months if you use yearly average for basis of calculation. Have you allowed for inverter, transmission and distribution losses (Loss of existing system in in the 14-18% range) Current electrical demand does not include conversion of all US vehicles to EV which is supposed to take place between 2025 and 2050. It appears you assume that all battery power consumption for one day, will be recouped when the sun starts shining the following day, but the following day, the generation must meet both demand and battery charging need... what if you have two or three consecutive cloudy days?

1blueheron is offline  
Old 04-26-2019, 05:49 AM
  #4  
Senior Member
 
crazybeard's Avatar
 
Join Date: Aug 2013
Location: Tropico
Posts: 5,807
Default

This is a very complex calculation and a lot depends on the system capabilities and distribution.

Lets look at it it from a simple math perspective though.

US consumed 3.95 trillion kWh in 18.

That comes out to a system that will need to output approx 18809523809 watts it located in AZ.

that would be 53741496 350w panels. Let's say they are 77x39in, roughly 20.9sq ft each.

that only just over a 40sq mile footprint for the array. surprizingly small.

they would need less of the array can adjust to the sun path dynamically.

Im not sure how to calculate the battery bank physical size, but let's give it a shot.

the average us per home Daily consumption kWh is about 29. I'm going to make an assumption that 60% of that is used when solar is not in use, so we need a daily capacity of about 17.5kWh.

each Tesla powerpack is rated to 210kWh meaning we need about 376190477 of them at 11.6sq ft ea. That's about 150.5 sq miles

now if they stack them 2 high, that would be 75.25sq miles. If they put solar above them, it could be doable and leaves 25sq miles for space between things and other equipment, extra capacity and expansion.

so yea it's probably doable in that area.



crazybeard is offline  
Old 04-26-2019, 05:53 AM
  #5  
Senior Member
 
Join Date: May 2014
Location: Wendell, NC
Posts: 2,190
Default

Considering he can't get his Teslas to stop catching on fire I'm not so sure I'd trust a 1 mile x 1 mile Tesla battery farm.
catmandoo1 likes this.
NCSUboater is offline  
Old 04-26-2019, 05:53 AM
  #6  
Senior Member
 
Join Date: Oct 2015
Posts: 995
Default

Like I said, very rough back of a soggy napkin...

figure was 3.95 trillion kilowatt hours/yr, or roughly 82,200,000kwh per day....

250w panel mounted properly in a decent location will give you plenty more than 1KWh/day.... I called it 1KWh/d to cover losses, the 2 sq meters per panel is quite generous (allowing room between for cleaning/maintenance)
the 80gwh battery is only adequate for usage smoothing over a single day.

your queries about winter demand etc is precisely why I added “triple the panels and increase the batteries by at least 5”
sullivan504 likes this.
mickt243 is offline  
Old 04-26-2019, 05:57 AM
  #7  
Senior Member
Thread Starter
 
Join Date: Jun 2014
Location: Virginia
Posts: 1,761
Default

Originally Posted by crazybeard View Post
This is a very complex calculation and a lot depends on the system capabilities and distribution.

Lets look at it it from a simple math perspective though.

US consumed 3.95 trillion kWh in 18.

That comes out to a system that will need to output approx 18809523809 watts it located in AZ.

that would be 53741496 350w panels. Let's say they are 77x39in, roughly 20.9sq ft each.

that only just over a 40sq mile footprint for the array. surprizingly small.

they would need less of the array can adjust to the sun path dynamically.

Im not sure how to calculate the battery bank physical size, but let's give it a shot.

the average us per home Daily consumption kWh is about 29. I'm going to make an assumption that 60% of that is used when solar is not in use, so we need a daily capacity of about 17.5kWh.

each Tesla powerpack is rated to 210kWh meaning we need about 376190477 of them at 11.6sq ft ea. That's about 150.5 sq miles

now if they stack them 2 high, that would be 75.25sq miles. If they put solar above them, it could be doable and leaves 25sq miles for space between things and other equipment, extra capacity and expansion.

so yea it's probably doable in that area.



I see the same errors here as with micks calcs. You can't use Average consumption and Average output if you are going to meet peak demands. This needs to be designed as an "off grid" PV system. Assume you have no generation capacity other than PV. If the sun doesn't shine for a couple days, you have total blackout. Food starts rotting, vehicles don't move, people are dying. You will need batteries capable of at least 48hrs. runtime, and PV generation capable of simultaneously recharging the batteries, and meeting peak demand of the grid plus taking into account conversion, transmission and distribution losses.

This is going to get tricky.
1blueheron is offline  
Old 04-26-2019, 06:01 AM
  #8  
Senior Member
 
Join Date: Oct 2015
Posts: 995
Default

Originally Posted by 1blueheron View Post
I see the same errors here as with micks calcs. You can't use Average consumption and Average output if you are going to meet peak demands. This needs to be designed as an "off grid" PV system. Assume you have no generation capacity other than PV. If the sun doesn't shine for a couple days, you have total blackout. Food starts rotting, vehicles don't move, people are dying. You will need batteries capable of at least 48hrs. runtime, and PV generation capable of simultaneously recharging the batteries, and meeting peak demand of the grid plus taking into account conversion, transmission and distribution losses.

This is going to get tricky.
Average/Average/Peak - precisely what the main function of the battery is, the battery drives the loads, the solar panels are there to charge the batteries
SOSC likes this.
mickt243 is offline  
Old 04-26-2019, 06:03 AM
  #9  
Senior Member
 
crazybeard's Avatar
 
Join Date: Aug 2013
Location: Tropico
Posts: 5,807
Default

Originally Posted by 1blueheron View Post
I see the same errors here as with micks calcs. You can't use Average consumption and Average output if you are going to meet peak demands. This needs to be designed as an "off grid" PV system. Assume you have no generation capacity other than PV. If the sun doesn't shine for a couple days, you have total blackout. Food starts rotting, vehicles don't move, people are dying. You will need batteries capable of at least 48hrs. runtime, and PV generation capable of simultaneously recharging the batteries, and meeting peak demand of the grid plus taking into account conversion, transmission and distribution losses.

This is going to get tricky.
i don't know the answer because I don't have those numbers. Would need the grid consumption details we don't have access to. I'm just doing math from a simple perspective.

You could use double the batteries from my example and stack them 4 high and make the array double or triple.

Tesla has done projects where they have access to a lot more use data plus I'm sure he can get a lot more just by asking during negotiations or other discussions.

I don't like the guy but I figure the smart people working for him are pretty good at what they do. I wouldn't be surprised if that is a fairly accurate statement on footprint.
sullivan504 likes this.
crazybeard is offline  
Old 04-26-2019, 06:03 AM
  #10  
Senior Member
Thread Starter
 
Join Date: Jun 2014
Location: Virginia
Posts: 1,761
Default

Originally Posted by mickt243 View Post
Like I said, very rough back of a soggy napkin...

figure was 3.95 trillion kilowatt hours/yr, or roughly 82,200,000kwh per day....

250w panel mounted properly in a decent location will give you plenty more than 1KWh/day.... I called it 1KWh/d to cover losses, the 2 sq meters per panel is quite generous (allowing room between for cleaning/maintenance)
the 80gwh battery is only adequate for usage smoothing over a single day.

your queries about winter demand etc is precisely why I added “triple the panels and increase the batteries by at least 5”
Noted. Tripling the number of panels in the same geographic region, subject to the same weather issues is a non-starter. 3 X 0= 0. You have to mitigate by distributing the generation facility across geographic regions. It has to be a "smart grid" with each array capable of backing up any other array. No single points of failure.
1blueheron is offline  
Old 04-26-2019, 06:10 AM
  #11  
Senior Member
 
Join Date: Oct 2015
Posts: 995
Default

Originally Posted by 1blueheron View Post
Noted. Tripling the number of panels in the same geographic region, subject to the same weather issues is a non-starter. 3 X 0= 0. You have to mitigate by distributing the generation facility across geographic regions. It has to be a "smart grid" with each array capable of backing up any other array. No single points of failure.
I never said all the panels had to be in the same spot.... I’d suggest a minimum of 3 locations, one on the west coast, one on the east and one roughly central.


better yet, put panels on every roof.....
mickt243 is offline  
Old 04-26-2019, 06:10 AM
  #12  
Senior Member
Thread Starter
 
Join Date: Jun 2014
Location: Virginia
Posts: 1,761
Default

Originally Posted by crazybeard View Post


i don't know the answer because I don't have those numbers. Would need the grid consumption details we don't have access to. I'm just doing math from a simple perspective.

You could use double the batteries from my example and stack them 4 high and make the array double or triple.

Tesla has done projects where they have access to a lot more use data plus I'm sure he can get a lot more just by asking during negotiations or other discussions.

I don't like the guy but I figure the smart people working for him are pretty good at what they do. I wouldn't be surprised if that is a fairly accurate statement on footprint.
Sure, the statement he makes is accurate in a perfect on paper world, just like his delivery and production dates are perfect on paper. However, common sense should tell us that you can't power Hawaii from Arizona. For that matter, you can't reasonably power the whole US from Arizona either. You have a single point of failure. Nobody of any intelligence would rely on a single central PV array for 100% of our electrical generation needs. This doesn't even take into account the implications of national security etc. It must be distributed and be redundant. N+1 or 2N at minimum and in multiple geographic areas.
1blueheron is offline  
Old 04-26-2019, 06:12 AM
  #13  
THT Sponsor
THT sponsor
 
Join Date: Nov 2018
Location: Melbourne, FL
Posts: 93
Default

I'm wondering what the environmental cost would be for producing all of those solar panels and batteries. I know the production of solar panels is mostly done in China because it uses some nasty chemicals and I've read about how massive the strip mining is to make the batteries. I'm sure we can get better in the future, but for now it seems like we're just moving or exporting our pollution.

It will be interesting to see how the grid handles the trend towards electric vehicles. I guess that's why Musk was working on the solar roof shingles.
NereusFishing is offline  
Old 04-26-2019, 06:16 AM
  #14  
Admirals Club Admiral's Club Member
 
Join Date: Nov 2008
Location: Somewhere in the middle of Michigan
Posts: 10,404
Default

Originally Posted by 1blueheron View Post
What figure are you using for "current US electrical demand"? Your system won't meet demands in the winter months if you use yearly average for basis of calculation. Have you allowed for inverter, transmission and distribution losses (Loss of existing system in in the 14-18% range) Current electrical demand does not include conversion of all US vehicles to EV which is supposed to take place between 2025 and 2050. It appears you assume that all battery power consumption for one day, will be recouped when the sun starts shining the following day, but the following day, the generation must meet both demand and battery charging need... what if you have two or three consecutive cloudy days?
Speaking of problems with the math, the losses in the existing system are smaller than the losses would be if the solar farm is in Arizona. The existing system has generating facilities scattered all over the nation. The mean distance from producer to consumer is shorter than in Musk's vision of all the production in Arizona and distribution to the 48 states. . As the mean distance goes up, so will the transmission losses.

Might be enlightening to calculate how many transmission lines would need to be run to the single-point source in Arizona to feed the rest of the nation. Might not even be possible to fit all the lines into the available space around a single 20 sq mi solar farm.

Let's also note that a single well-placed missle takes out power to the entire nation.

So, Musk has an interesting vision, but that's probably as far as it will go.
yarcraft91 is offline  
Old 04-26-2019, 06:18 AM
  #15  
Senior Member
 
crazybeard's Avatar
 
Join Date: Aug 2013
Location: Tropico
Posts: 5,807
Default

Originally Posted by 1blueheron View Post
Sure, the statement he makes is accurate in a perfect on paper world, just like his delivery and production dates are perfect on paper. However, common sense should tell us that you can't power Hawaii from Arizona. For that matter, you can't reasonably power the whole US from Arizona either. You have a single point of failure. Nobody of any intelligence would rely on a single central PV array for 100% of our electrical generation needs. This doesn't even take into account the implications of national security etc. It must be distributed and be redundant. N+1 or 2N at minimum and in multiple geographic areas.
his statement is just about capacity and to make a point.

Your points are valid too.

IMO a completely distributed system where each building produces their own would be best, but slow and expensive to scale along with the corporate and political implications in addition more costly due to inefficiency from less insolation in various areas. But perhaps the reduction or elimination of transmission lines and the benefits from that may be significant.
crazybeard is offline  
Old 04-26-2019, 06:27 AM
  #16  
dev
Admirals Club Admiral's Club Member
 
dev's Avatar
 
Join Date: Mar 2017
Location: St. Pete
Posts: 499
Default

It's a thought exercise, not a proposal for an actual power grid.
You would power Hawaii from power generation sources located on the islands, same as today.
Batteries would be more practical in a distributed fashion, not centrally located. Medium/small battery at each house, maybe larger regionalized battery banks.

I think the major take-away is that relatively speaking it would not take much area to create a solar grid that could supply the *average* power demands of the US. That does not mean solar is the 100% perfect choice, but that it is a very viable option for supplying a significant part of the power needs for the US.

There are lots of places in the southern US that would make good locations for solar arrays. There are even places in the north that would, it would take more panels, but not to the degree that there is not enough space to locate them.
mattttt25 likes this.
dev is offline  
Old 04-26-2019, 06:30 AM
  #17  
Senior Member
 
Join Date: Oct 2015
Posts: 995
Default

Further rough numbers (based on the 10kw system I had installed recently), and assuming $1k for a 1KWh battery....and sticking to 250w panels, because they are plentiful and cheap....

lets break the system down, so as to not have all ~80GWh/d production in a single spot (the grid would melt!)

82,200,000 - for simplicity and redundancy, let’s round up to 100 million units, of a single 250w panel, micro inverter and 1KWh battery, will cost somewhere in the vicinity of $1500us per unit.

total installed cost of $150,000,000,000. Not that bad really....... again, triple it fro weather resistance etc.
half a trillion...... whats the replacement cost of a power Plant? Fuel costs?

how many houses in the us? 10 units per house on 30 million houses, you’d get a lot mor on factories and warehouses..... so a low install cost per house/building

mickt243 is offline  
Old 04-26-2019, 06:52 AM
  #18  
Senior Member
 
Join Date: Feb 2017
Posts: 1,823
Default

Originally Posted by crazybeard View Post
US consumed 3.95 trillion kWh in 18.

That comes out to a system that will need to output approx 18809523809 watts it located in AZ.
Not sure how you got those numbers, but assuming the 3.95 trillion kWh is correct, then that comes out to a continuous need for 450 BILLION watts to be generated in AZ. If you figure on average you'd get 1kw per square meter from the sun and can only use that 6-8 hrs/day, that's a total (average) of 7kWh/day/sqmeter at 100% efficiency of a solar cell. If you are generous and estimate future solar cell efficiency at 20% and tightly space them, that means you need 2,821 billion square meters of solar cells to meet the 3.95 trillion kWh/yr.

That comes out to about 1,069 sq miles of surface for solar cells.

Now, as pointed out previously, you sometimes have to charge your battery packs (or whatever storage method you use) WHILE you're still providing full load to the grid, which means you need to DOUBLE or TRIPLE the total generation based on how long you can take to recharge your storage and that's only if the storage is good for one or two days of "blackout". You are already using the capacity to recharge the storage for each overnight scenario which is a known issue.
km1125 is offline  
Old 04-26-2019, 06:59 AM
  #19  
Admirals Club Admiral's Club Member
 
Join Date: Nov 2009
Location: Tarpon Springs, FL
Posts: 7,413
Default

They just put in a large solar field off I-75 in North Florida. The panel density is no where near what you guys are figuring. Plus I would be concerned about transmission line losses if all of the production was in one place. As the technology improves medium sized fields in a lot of locations will be a viable option. As for the batteries that's a whole nother ball game.
mikefloyd is offline  
Old 04-26-2019, 07:08 AM
  #20  
Senior Member
 
Join Date: Oct 2015
Posts: 995
Default

Originally Posted by km1125 View Post
Not sure how you got those numbers, but assuming the 3.95 trillion kWh is correct, then that comes out to a continuous need for 450 BILLION watts to be generated in AZ. If you figure on average you'd get 1kw per square meter from the sun and can only use that 6-8 hrs/day, that's a total (average) of 7kWh/day/sqmeter at 100% efficiency of a solar cell. If you are generous and estimate future solar cell efficiency at 20% and tightly space them, that means you need 2,821 billion square meters of solar cells to meet the 3.95 trillion kWh/yr.

That comes out to about 1,069 sq miles of surface for solar cells.

Now, as pointed out previously, you sometimes have to charge your battery packs (or whatever storage method you use) WHILE you're still providing full load to the grid, which means you need to DOUBLE or TRIPLE the total generation based on how long you can take to recharge your storage and that's only if the storage is good for one or two days of "blackout". You are already using the capacity to recharge the storage for each overnight scenario which is a known issue.
I think you’re off by a factor of 1000...

assuming that a trillion is 1,000,000,000,000


your 450gigawatt generator puts out 10,800 gigawatt hours per day

or 10.8 terrawatt hours per day, or about 2.5 times the us annual consumption per day.

mickt243 is offline  

Thread Tools
Search this Thread