Can I do this in a few parts? Tom Swift was good enough to link some EIA numbers I posted to him here
Subject 54254
The easy answer to your question is it varies. But if you will allow me I'd like to back up a bit and talk more generally before talking about percentages that these costs make up.
As you know, for the most part, electricity cannot be conveniently stored: So instantaneous supply must balance instantaneous demand. And the demand for electrical energy in not only variable, but quite nonlinear, especially because of the HVAC demands of residential and commercial customers. For any given system or control area -- if you took the load seen in each hour of the year -- 8760 -- or 8784 for leap years and ranked them descending you would have what we call a load duration curve. For most systems the peak load can be half again or even more than the average load -- owing to that HVAC sensitivity. But the peaked portion of the curve is somewhat narrow. At the other end of the curve -- near the lowest loads the difference with the annual average is smaller. The middle or fat of the curve is rather gently sloping.
Why so much detail about the shape of the load-duration curve?? Well this shape dictates the mix of generating capacity that utilities use to serve the load that they see throughout the year. For the load which persists throughout the year, fixed costs can be spread to all the hours so that the total cost per hour per increment of energy generated is lowered -- i.e. a plant that has higher fixed costs, but the lower or lowest operating costs can be used to serve the "base" load. In the peaked portion of the load curve -- that load doesn't have to be served for very long so you want the reverse -- you can tolerate higher operating costs (small increment of the total energy generated throughout the year), but you want the fixed costs to be low, because you are spreading them to so few hours. Of course, there is that sloping portion in between -- which is "intermediate" load. To serve this increment of load you can tolerate a little higher fixed costs than the baseload and higher operating costs, since the energy, though not negligible, is not near the amount in the "baseload" portion of the curve.
At the bottom of the load duration curve you find baseload units such as coal or nuclear plants. I'm not terribly familiar with nukes, as we don't have them, but know a bit about the costs associated with coal.
In the middle, typically are combined cycle (CCs) gas plants and/or less efficient coal plants. A combine cycle plant is a gas combustion turbine that has exhaust gases fed to a HRSG (heat recovery steam generator) that can use exhaust heat to produce steam to drive a second turbine -- some more advanced designs will use a single shaft.
At the top are simply cycle combustion turbines, CTs.
To give you an idea on the differences in fuel costs --
The fuel alone for a coal plant can cost anywhere from
1.00 to 1.80 mmBtu depending on how far it has to be transported. A lot of the price consists of transporation.
The fuel for either a CT or CC if nat gas is used has been bouncing around between the high 4.XX and 5.XX.
A typical heat rate for a coal plant (this is the conversion ratio from fuel cost to generation cost) is around 10 - 10.5. For a simple cycle combustin turbine the number is around 11-13. For a combined cycle plant (fairly efficient machines) the number can be from 7 - 8.8 or so.
So you see that for the fuel cost (given the narrow nat gas prices above)
Coal could range from $10-20/MWh
CT could range from $44 - $65/MWh
CC could range from $28 - $44/MWh
Fuel of course is just one part of the operating cost
Usually O&M is split into fixed and variable
For CC's I've seen as low as $1.50-$2.00/MWh classified as variable. With the machines running from about 30%-45% of the time.
It's a bit weird, in the classification, but CTs usually
around $5 - $6/MWh -- makes some sense when you consider they might run only 5% of the time -- spread the cost to fewer hours, etc.
I've argued with folks about how much is truly variable with respect to coal -- but we are currently using around
1.25 /MWh as truly variable -- but there is a large fixed component for coal as you could imagine.
I don't know what IOUs do for depreciation -- they may have lives dictated by regulation, but we typically use straight line depreciation and 30 years.
You can get some idea of the capital costs by looking at that EIA chart I mentioned at the first.
A typical CT will be 100 MW -- some smaller machines are in the 40-50 size.
A combine cycle might be 250 or 500 MW depending on the configuration.
Coal to be economic, these plants have to be built in sizes of 600 MW or more -- there are smaller plants, but the economies of scale push you to partner if you don't have that much demand yourself.
I'm not sure what you mean by profit margin -- are you asking what profit margin would be required to put the money at risk? We are currently looking at a coal plant in the next decade (or a share of one) -- price tag about 1.1 B -- for us, with a captive customer base - it is not about profit, but finding the cheapest means for our collection of resources to provides energy at the lowest overall cost to our customers. There are off system sales -- and we would take into account -- though very conservatively -- what profits we would make on those sales -- but only to the extent that those sale would lower the overall costs to our customers. |
