Refining

Refining crude oil into gasoline takes energy. But how much?

From this page: http://tonto.eia.doe.gov/dnav/pet/pet_pnp_capfuel_dcu_nus_a.htm
We can see that in 2008 american refineries used:
711 billion cubic feet of natural gas
43000 tons of coal
42.7 billion kWhr of electricity purchased from the grid
327 million barrels of oil products ( non-gasoline refining leftovers )
98.8 billion pounds of steam purchased

From this source we find the conversion efficiency of electric power generating stations ( in 2005/2006) . www.npc.org/Study_Topic_Papers/4-DTG-ElectricEfficiency.pdf
At that time the average BTU/kWh for US coal was 10,410
The top 20 US coal plants averaged 9400 BTU/kWhr
The best available technology for a new coal plant was 7757 BTU/kWhr
The average US NG plant was 7920 BTU/kWhr
A typical NGCC ( natural gas combined cycle ) plant is 6800 BTU/kWhr
And the best available NGCC technology is 6333 BTU/kWhr

Back to our consumpton numbers… if instead of using that energy to refine crude oil into gasoline, we used it to generate electricity, how much would we get?

The natural gas could generate 90 billion kWhr with average gas plants.
And 107 billlion kWhr using the best NGCC plants.
The coal can produce 107 billion kWhr using average coal plants, or 139 billion kWhr using the most advanced coal plants.
I couldnt find data on electrical generation from petroleum products, so I’m going to take the average of coal and NG ( the average plants ). Thus the 327 million barrels of oil products can produce 235 billion kWhr of electricity.
Lastly the energy used to generate the steam could instead generate 13 billion kWhr of electricity.
Add all that up and you get a low value of 488 billion kWhr and a high value of 533 billion kWhr.

In 2008 the total number of vehicle miles driven was 2921 billion.
The Tesla Roadster is about 88% efficient charging and there is about 8% transmission loss per 500 miles. At 80% charging efficiency and 4 miles per kWhr, the 488 billion kWhr would move EVs a total of 1561 billion miles.

Yes we could supply the energy we need to drive EVs more than 50% of the miles we drive in a year from repurposing the energy we use just to refine gasoline.

This does not even cover the energy that we use to extract crude oil and transport gasoline to the consumer.

Wind Energy Growth, Coal Energy Decline

Electricity production share from wind has been growing in the U.S. and electricity production share from coal has been declining over the last few years.
Thats good news!
In 1997 coal was 52.8% of U.S. production.
2004 coal was 49.8% ( went under 50% for the first time )
2006 coal was 49%
2008 coal was still 48.5%
The 12 months period July 2008 to June 2009 coal was 46.6% ( call this period 2008-09 )
From 1997 to 2008-09 electricity from non-hydro renewables is up from 2.1% to 3.2%, mostly from growth in wind power ( this growth has actually all happened in the last 4 years ) The growth of wind can’t explain all of the decline in coal.
Turns out that natural gas is way up from 13.7% in 1999 to 22.1% in 2008-09.

Why is coal down so much in the last 12 months?
Overall electricity use is down significantly in the last 9 months.
But that decline is all in one sector – residential electricity use is down 1% and commercial electricity use is down 2% and industrial electricity use is down 13%.

Natural gas is burst electricity, it’s what companies spin up for the peak of the day when we all turn on our lights and AC and TVs – anything that varies during the day – residential and commercial.
Industrial use has more round the clock use, so that means coal and nuclear. Between coal and nuclear, you turn off coal first.

Okay but why is electricity from coal ( as a percentage of total production ) down over the long term?
Going back to look at the last 10 years, its pretty obvious that the residential use has been growing while industrial use is pretty flat.
That explains why natural gas has been gaining share on coal over that time.

Wind power is growing, but its still a minor player.

more electricity production trivia

I found a chart from this ancient BPA document to illustrate the hourly demand curve in the pacific northwest.

Obviously the pacific northwest is a lot different from California. We have a lot of electric heat that we use in the winter, and there is very little residential air conditioning.
I would think that one big reason our power is so cheap here compared to California is that we dont have to build a lot of generation we dont need for those really high summer peaks driven by air conditioning.

more electricity production trivia

I finally found some pretty charts to help visualize the demand curve during the day.
Here is one from California from today:

The temperature range in Los Angeles today was 66F – 88F and San Jose was 61F – 95F ( peaked at 3pm )

I found it on this webpage which is from the California power consortium. There is plenty of historical data here as well that I will try to dig into soon.

I also found this data from Ontario Canada. The demand curve is significantly different shape, probably because of a different mix between AC and lighting.
It was a mild day today in Ontario, temperatures from 9C – 20C. ( 48F – 68F )

So my chart that I made yesterday was a little off. The peak consumption in California is at 4PM and there is a long tail with continuing high demand late into the evening. The difference between the 2 charts seems to strongly indicate that the AC load of a hot day pushes the peak higher centered around 3-4PM, otherwise it is flatter throughout the work hours. Hopefully I can find some historical data for a really hot day from both places.

It looks like in California there is a lot of surplus power in the morning all the way up til the early afternoon. If you need to charge your car after driving to work, you’ll want to do it in the morning between 7AM and 2PM or so, and not in the late afternoon. Optimal charge time overnight looks like between 1AM and 6AM, but any time from about 10PM til 7AM seems fine.

electricity production trivia

I was curious about some details of electricity production in the US, so I dug up some data.
The total production over the year in 2007 and 2008 was about 4,156 billion kWhr and 4,110 billion kWhr. ( Production was actually down in 2008 slightly from 2007, mostly in the last half of the year )
The maximum instantaneous production in the US in 2007 was about 1 billion kW.
There are 8760 hours in the year, so over the course of the year we run our generators at about 47% capacity.

Why is it so low? Because max capacity is needed for only a brief period in the middle of the day in the highest demand days in august.

Over the year the big 3 of energy production are used at a fraction of their capacity. All the other sources are around 10% of total production ( hydro, wind, geothermal, petroleum products, etc ) so we’ll ignore them for now.
Coal runs at 73%, Natural gas is 25% and nuclear is at 92% of generating capacity.
Why? Nuclear has the highest up front cost and the lowest fuel cost, and natural gas has the lowest up front and highest fuel cost with coal in the middle.

July and August are the months with the highest electriciy demand. During August 2007 we used 58% of capacity.
April is the month with the lowest demand, during April we average using 42% of capacity.

In August, coal averages 83%, natural gas 42% and nuclear 99% of capacity.
In April, coal averages 64%, natural gas 21%, and nuclear 78% of capacity.
Why do we burn natural gas at all if we have unused coal capacity? Because coal and nuclear are slow to start and stop, you dont cycle them on and off during the day, they take hours to heat up and cool down. You only stop a coal fired boiler if you are going to idle it for a lot longer than one day.
Natural gas is used to fill in the peak of the day, while the coal and nukes probably run all 24 hours.

I have been unable to find real data on the production/consumption curve during the day.
So picking August when we run 58% capacity here are 3 possible curves for an August day.

power consumption curve estimates

The green line would be constant 58% through the day. We know this doesnt happen because nobody would have built that other 42% if it was so.
The peak happens sometime early afternoon, and when it gets too close to 100% or exceeds it you get brownouts. All the lines have the same area under the curve, I have no idea which one is closest to reality. The blue line bottoms out at 37% which is where you would be if you only ran nuclear and coal and everything else shut down ( which is actually unlikely: hydro, geothermal, wind and others dont shut down either )
The point of that? To show that in the hours outside of about a 4 – 6 hour range in the middle of the day there is tons of excess capacity.