Friday, August 19, 2005
A bite-sized cogeneration example
Continuing a thought started elsewhere, I thought I'd blog something about cogeneration and plug-in hybrids. (Note that this is a quickie and not thoroughly cross-checked, there could be errors in the calculations - caveat reader.)
The plug-in hybrid concept is touted for its potential to eliminate petroleum, but there are valid questions about how quickly we could increase generation to supply them. This led me to ask, what if we used petroleum to supply the energy? Better yet, if it was used in cogenerators?
Assume for the moment that refined petroleum (sulfur and heavy metals reduced to acceptable levels) costs $12/GJ (crude currently runs about $11/GJ). Company X switches its process heat system from natural gas at $10.00/MMBTU ($9.48/GJ) and 10% losses to a cogenerator system co-firing natural gas and fuel oil at 50% efficiency and 10% losses. The company buys 1.25 GJ of oil to supplement each GJ of gas they formerly used, sells the electricity for $0.08/kWh ($22.22/GJ), loses 10% and uses 40%; their fuel costs go from $10.54/GJ delivered to the process to $24.48, offset by $27.78 in electricity sales. The gross cost falls from $10.54 to $4.98/GJ of process heat; the cost of heat is cut in half.
The 2.25 GJ of fuel produces 1.125 GJ of electricity. This travels over the grid, losing perhaps 10% en route to the customer; 1.01 GJ is delivered. It is purchased by a GO-HEV owner whose vehicle uses 350 WH/mile (1.26 MJ/mile) at the charger. The vehicle achieves 794 miles per GJ at the wall, or 357 miles per GJ of fuel delivered to the cogenerator; if the owner pays $0.12/kWh, the per-mile cost for electricity is 4.2 cents.
Last, assume that the electricity delivered to the GO-HEV replaces gasoline at the rate of 1/35 gallon per mile.
A barrel of crude oil is 42 gallons. If we assume that the refined fuel oil has the same 6.1 GJ/bbl energy value as crude (probably not too far off), here's what the cogenerating system would accomplish with each barrel of fuel oil:- Reduce process heat costs by $27.08/bbl ($5.55/GJ process heat savings / 1.25 GJ oil per GJ heat * 6.1 GJ/bbl).
- Provide 1525 kWh of electricity to the utility, of which 1372 kWh is delivered at a gross profit of $42.70/bbl.
- Allow the driver to cover 3921 miles at an effective economy of 93.4 MPG (52.4 MPG per gallon-equivalent of all fuel used).
- Displaces 112 gallons of gasoline.
- Save the driver $132.21 (3.37¢/mi) in energy costs compared to gasoline at $2.65/gallon.
- Save a total of 70 gallons of fuel (42 gallons used vs. 112 displaced).
Gross cost savings and increased profits: $223.71/bbl. We have a number of policies regarding electric generation (existing regulations appears to discourage cogeneration), vehicles (air-quality policy encourages the small improvements produced by ethanol but not the much larger improvements possible with partial ZEV operation) all of which stand in the way of accomplishing this. If we fail to examine these policies and make appropriate changes ASAP, we are fools. (Posting this entry was an exercise in hair-tearing; see Stupid Blogspot tricks.) UPDATE 2005-Aug-29: I think I might have the figures correct this time. As confirmation, here is a before-and-after table of energy expenditures and disposition for comparison purposes (not including losses in electric transmission or profits to the provider of electric transmission, which could account for the stray twenty bucks between this table and the list above and which I would have put in if I could think of a good way to represent it):
| System | Fuel | bbl (equiv.) | Fuel cost, $ | Heat, GJ | VMT | Net fuel cost, $ |
| Industry | Nat. gas | 1.0 | 57.83 | 5.49 | 57.83 | |
| Vehicle | Gasoline | 2.67 | 296.91 | 3921 | 296.91 | |
| 3.67 | 354.74 | 5.49 | 3921 | 354.74 | ||
| Cogen | Nat. gas | 1.0 | 57.83 | 57.83 | ||
| Cogen | Fuel oil | 1.25 | 91.50 | 91.50 | ||
| 2.25 | 149.33 | 5.49 | 3921 | 149.33 | ||
Comments:
<< Home
(If you made an egregious error in your comment, your solution is to click the delete icon and enter a corrected version. Blogger provides no facility for a blog owner to edit comments; it's delete or leave as-is.)
I have a couple questions about the increased demand for electricity.
Eyeballing this, I gather there'll be crudely somewhere around a 50% (plus or minus a lot) increase in overall demand for electricity (assuming transportation and electricity each are roughly equal consumption of energy and half of the transportation consumption is for small automobiles).
Second as you mention, it appears with a little cleverness or market incentives, the overall demand curve for electricity will flatten out.
Due either of these have some implication for electricity production or transmission? Obviously, we'll need to build up the infrastructure, but will this encourage any sort of new technologies or practices in the electricity industry?
Eyeballing this, I gather there'll be crudely somewhere around a 50% (plus or minus a lot) increase in overall demand for electricity (assuming transportation and electricity each are roughly equal consumption of energy and half of the transportation consumption is for small automobiles).
Second as you mention, it appears with a little cleverness or market incentives, the overall demand curve for electricity will flatten out.
Due either of these have some implication for electricity production or transmission? Obviously, we'll need to build up the infrastructure, but will this encourage any sort of new technologies or practices in the electricity industry?
Electricity infrastructure is not what's limiting plug-in hybrid adoption, and lack of demand for electricity is certainly not limiting co-generation.
Europe's got plenty of incentives for co-generation and sky high gasoline taxes.
The real limiting factors are to be found in the nexus of investment cost / performance / convenience.
Co-generation is used where there is a large constant demand for heat, be it in oil refineries or in oil sands production or for ethanol production from sugar cane.
When the demand for heat is not constant or large, the investment cost for the generating unit beomes quite substantial, and/or the efficiency goes down, and/or "convenience" becomes an issue.
For example, the most efficient way to heat homes is to have large blocks of flat supplied with district heat from a co-generator, with the electrical output used to drive heat pumps.
That way each kWh of natural gas can provide 3 kWh of heat (even better, if cooking, drying, hot water and the like are met that way, a surprisingly large fraction of electricity consumption is for resistance heating).
Not only that, if it's tightly clustered flats, the requirement in kWh of heat per m2 will go down massively as well.
So that in sum total, the requirement for nat gas can easily be brought down by a factor ten when compared to natural gas heated detached family homes providing the same floor space, and that without insulating any more.
But many people want to live in detached homes with a large garden. Co-generation then becomes astronomically expensive per kWh of natural gas saved, not least because the small units available for family home sized demand have quite low efficiency for electricity generation (Whispergen being the worst I've come across).
We've discussed batteries to death, yes, I know, my main problem with goingreen is that I am not in London. But my real main problem is that I just cannot buy a suitable electric vehicle at a vaguely reasonable price.
If it was such a good deal, why on earth is nobody offering one?
Europe's got plenty of incentives for co-generation and sky high gasoline taxes.
The real limiting factors are to be found in the nexus of investment cost / performance / convenience.
Co-generation is used where there is a large constant demand for heat, be it in oil refineries or in oil sands production or for ethanol production from sugar cane.
When the demand for heat is not constant or large, the investment cost for the generating unit beomes quite substantial, and/or the efficiency goes down, and/or "convenience" becomes an issue.
For example, the most efficient way to heat homes is to have large blocks of flat supplied with district heat from a co-generator, with the electrical output used to drive heat pumps.
That way each kWh of natural gas can provide 3 kWh of heat (even better, if cooking, drying, hot water and the like are met that way, a surprisingly large fraction of electricity consumption is for resistance heating).
Not only that, if it's tightly clustered flats, the requirement in kWh of heat per m2 will go down massively as well.
So that in sum total, the requirement for nat gas can easily be brought down by a factor ten when compared to natural gas heated detached family homes providing the same floor space, and that without insulating any more.
But many people want to live in detached homes with a large garden. Co-generation then becomes astronomically expensive per kWh of natural gas saved, not least because the small units available for family home sized demand have quite low efficiency for electricity generation (Whispergen being the worst I've come across).
We've discussed batteries to death, yes, I know, my main problem with goingreen is that I am not in London. But my real main problem is that I just cannot buy a suitable electric vehicle at a vaguely reasonable price.
If it was such a good deal, why on earth is nobody offering one?
OK, that's not an electric car, but a micro CHP plant.
I couldn't find a price, but I am aware of the options available in Germany:
http://www.energienetz.de/pre_cat_41-id_155-subid_1059-subsubid_1074__.html
5 kW electric, 12 kW thermal, electric efficiency 28%, 10% losses, remaining 62% thermal.
Cost: €18,000
They calculate:
Energy costs, 102,500 kWh of natural gas at 2.5 cents per kWh, ~€2,500 per year
Other costs ~€1,000 per year
Value of heat (62,500 kWh) at 3.7 cents per kWh ~€2,300
Value of electricity (most sold to the grid, as heat and electricity demand don't coincide) 27,500 kWh at 3 cents per kWh ~€800
Subsidies ~€1,900
Total surplus, ex capital cost, including subsidies: ~€1,300
Payback time at 6% interest: 30 years.
62,500 kWh of heat is a lot. The average living space per capita in Germany is 40 square metres, in the US it's about 70. For a 3 person household in the US that's about 200 square metres. New housing in Germany requires about 100 kWh/m2 of heat, assuming that applied to the US as well, we'd be talking 20,000 kWh for an average three family home, so the Dachs would be heavily oversized.
Are there smaller units?
Well, yes, eg Whispergen, with its grand electric efficiency of 10% ...
To cut a long story short, co-generation can be a great money maker, but there's a wide range of profitability depending principally on how big and steady the heat demand is, and those opportunities for co-gen not in use tend to, as you'd expect, to heavily cluster towards the uneconomic end of that spectrum.
I couldn't find a price, but I am aware of the options available in Germany:
http://www.energienetz.de/pre_cat_41-id_155-subid_1059-subsubid_1074__.html
5 kW electric, 12 kW thermal, electric efficiency 28%, 10% losses, remaining 62% thermal.
Cost: €18,000
They calculate:
Energy costs, 102,500 kWh of natural gas at 2.5 cents per kWh, ~€2,500 per year
Other costs ~€1,000 per year
Value of heat (62,500 kWh) at 3.7 cents per kWh ~€2,300
Value of electricity (most sold to the grid, as heat and electricity demand don't coincide) 27,500 kWh at 3 cents per kWh ~€800
Subsidies ~€1,900
Total surplus, ex capital cost, including subsidies: ~€1,300
Payback time at 6% interest: 30 years.
62,500 kWh of heat is a lot. The average living space per capita in Germany is 40 square metres, in the US it's about 70. For a 3 person household in the US that's about 200 square metres. New housing in Germany requires about 100 kWh/m2 of heat, assuming that applied to the US as well, we'd be talking 20,000 kWh for an average three family home, so the Dachs would be heavily oversized.
Are there smaller units?
Well, yes, eg Whispergen, with its grand electric efficiency of 10% ...
To cut a long story short, co-generation can be a great money maker, but there's a wide range of profitability depending principally on how big and steady the heat demand is, and those opportunities for co-gen not in use tend to, as you'd expect, to heavily cluster towards the uneconomic end of that spectrum.
Getting back to substantive responses:
Karl: Yes, total energy demand would rise substantially. (If most of the energy was sent through high-energy but lossy batteries like zinc-air, it might rise by almost 80%.)
The effect on generating capacity depends a lot on what kind of batteries are used and how much capacity is available. Vehicles such as GO-HEVs which use small packs will need them to be fully charged every night, so the demand level for each night will follow the previous day's driving. This requires dispatchable capacity, though DSM can allow changes in power level to be done over hours rather than minutes. This plays well to base-load powerplants (coal or nuclear).
If the batteries are something like externally-fuelled zinc-air cells, the energy supply stored outside the vehicle can be much larger than what's on board. This would allow demand to be tracked (or scheduled) days in advance, and resources like wind and solar could supply a lot of the energy. Shortfalls would appear over days and could be made up over days; you could wait for the next front to come by and let the wind farms top up your energy storage.
Even short-range GO-HEV systems which have vehicles connected to the grid for much of the day could allow DSM to eliminate a great deal of simple-cycle gas turbine capacity (fast load-following) in favor of combined-cycle gas turbines.
The changes in the system appear to be flattening of the daily load curve, a shift from peaking plants to more efficient and cheaper base-load plants, and in some scenarios (lotsa storage) you get the ability to make full use of very large amounts of non-schedulable generation. Transmission requirements from conventional plants would stay flat, transmission requirements from wind farms would increase.
Karl: Yes, total energy demand would rise substantially. (If most of the energy was sent through high-energy but lossy batteries like zinc-air, it might rise by almost 80%.)
The effect on generating capacity depends a lot on what kind of batteries are used and how much capacity is available. Vehicles such as GO-HEVs which use small packs will need them to be fully charged every night, so the demand level for each night will follow the previous day's driving. This requires dispatchable capacity, though DSM can allow changes in power level to be done over hours rather than minutes. This plays well to base-load powerplants (coal or nuclear).
If the batteries are something like externally-fuelled zinc-air cells, the energy supply stored outside the vehicle can be much larger than what's on board. This would allow demand to be tracked (or scheduled) days in advance, and resources like wind and solar could supply a lot of the energy. Shortfalls would appear over days and could be made up over days; you could wait for the next front to come by and let the wind farms top up your energy storage.
Even short-range GO-HEV systems which have vehicles connected to the grid for much of the day could allow DSM to eliminate a great deal of simple-cycle gas turbine capacity (fast load-following) in favor of combined-cycle gas turbines.
The changes in the system appear to be flattening of the daily load curve, a shift from peaking plants to more efficient and cheaper base-load plants, and in some scenarios (lotsa storage) you get the ability to make full use of very large amounts of non-schedulable generation. Transmission requirements from conventional plants would stay flat, transmission requirements from wind farms would increase.
Heiko writes:
"Co-generation is used where there is a large constant demand for heat..."
It appears that your addition of "large" and "constant" to the requirements is just that, yours. The Climate Energy unit is designed for a relatively small (13,800 BTU/hr) heat demand (with supplemental heat available as required) and cycles as appropriate.
"When the demand for heat is not constant or large, the investment cost for the generating unit beomes quite substantial"
USD 8000 total, USD 4000 premium over a conventional unit. I personally think this is awfully steep considering that it's more than the cost of the typical automobile drivetrain, but that may reflect low volume and stiff emissions limits on grid-connected stationary generators.
"That way each kWh of natural gas can provide 3 kWh of heat"
You're not going to get that kind of performance. Even with a solid-oxide fuel cell at 60% efficiency, you're only going to get 40% waste heat + 3x60% heat-pump output = 2.2 CoP. With an internal combustion engine at 30% efficiency, you'll peak out at maybe 1.7-1.8.
What the cogenerator/heat pump system can do is absorb a huge amount of intermittent electric supply, such as wind, solar or wave power. The cogenerators can be run just enough to make up the difference between the electric supply and the heat demand. Even 30% capacity factor on the other generation displaces a lot of fuel.
"Not only that, if it's tightly clustered flats, the requirement in kWh of heat per m2 will go down massively as well."
I'll let you do the redevelopment of areas of single-family homes to apartment and condominium blocks. I hope you have a lot of money to spend for zoning changes and legal disputes, plus the loss on the property values...
"the requirement for nat gas can easily be brought down by a factor ten"
Gas requirements could be rendered negligible if you made people live in little concrete-skinned, spray-foam domes. Unfortunately for your dream world, this would never happen easily; people will not abandon their space and creature comforts unless they are forced by necessity or at gunpoint.
You're a Brit, maybe you'd leave your home at the command of an energy czar. Americans would shoot the people who tried to force them, and would come up with alternatives if conventional energy supplies ran short.
"But many people want to live in detached homes with a large garden. Co-generation then becomes astronomically expensive per kWh of natural gas saved..."
The Climate Energy unit costs about $4000 more than a conventional heating plant and generates 1 kW when in operation. A $4000 premium amortized at 7% over 15 years is roughly $440/year. If it has an annual duty cycle of 45% (including water heating), it would generate 3942 kWh (worth about $670 at Californian rates of 17¢/kWh, perhaps more in a place like New York City) while the added fuel cost at $1.00/therm is about $135. Looks like the owner could clear $100/year before taxes, and given the US deductibility of mortgage interest it would be about $178 after taxes. If much of the electricity was used in an EV or GO-HEV its value (avoided cost of gasoline) would be upwards of 20¢/kWh.
The economics in your Energienetz example are determined by the sale price of the electricity, which is terribly low. Isn't the German government mandating much larger payments for wind and solar power? Regardless, if that electricity was used to power vehicles the avoided cost of fuel would be much greater than in the US. I'll let you work the example because I don't know the fuel costs.
"Co-generation is used where there is a large constant demand for heat..."
It appears that your addition of "large" and "constant" to the requirements is just that, yours. The Climate Energy unit is designed for a relatively small (13,800 BTU/hr) heat demand (with supplemental heat available as required) and cycles as appropriate.
"When the demand for heat is not constant or large, the investment cost for the generating unit beomes quite substantial"
USD 8000 total, USD 4000 premium over a conventional unit. I personally think this is awfully steep considering that it's more than the cost of the typical automobile drivetrain, but that may reflect low volume and stiff emissions limits on grid-connected stationary generators.
"That way each kWh of natural gas can provide 3 kWh of heat"
You're not going to get that kind of performance. Even with a solid-oxide fuel cell at 60% efficiency, you're only going to get 40% waste heat + 3x60% heat-pump output = 2.2 CoP. With an internal combustion engine at 30% efficiency, you'll peak out at maybe 1.7-1.8.
What the cogenerator/heat pump system can do is absorb a huge amount of intermittent electric supply, such as wind, solar or wave power. The cogenerators can be run just enough to make up the difference between the electric supply and the heat demand. Even 30% capacity factor on the other generation displaces a lot of fuel.
"Not only that, if it's tightly clustered flats, the requirement in kWh of heat per m2 will go down massively as well."
I'll let you do the redevelopment of areas of single-family homes to apartment and condominium blocks. I hope you have a lot of money to spend for zoning changes and legal disputes, plus the loss on the property values...
"the requirement for nat gas can easily be brought down by a factor ten"
Gas requirements could be rendered negligible if you made people live in little concrete-skinned, spray-foam domes. Unfortunately for your dream world, this would never happen easily; people will not abandon their space and creature comforts unless they are forced by necessity or at gunpoint.
You're a Brit, maybe you'd leave your home at the command of an energy czar. Americans would shoot the people who tried to force them, and would come up with alternatives if conventional energy supplies ran short.
"But many people want to live in detached homes with a large garden. Co-generation then becomes astronomically expensive per kWh of natural gas saved..."
The Climate Energy unit costs about $4000 more than a conventional heating plant and generates 1 kW when in operation. A $4000 premium amortized at 7% over 15 years is roughly $440/year. If it has an annual duty cycle of 45% (including water heating), it would generate 3942 kWh (worth about $670 at Californian rates of 17¢/kWh, perhaps more in a place like New York City) while the added fuel cost at $1.00/therm is about $135. Looks like the owner could clear $100/year before taxes, and given the US deductibility of mortgage interest it would be about $178 after taxes. If much of the electricity was used in an EV or GO-HEV its value (avoided cost of gasoline) would be upwards of 20¢/kWh.
The economics in your Energienetz example are determined by the sale price of the electricity, which is terribly low. Isn't the German government mandating much larger payments for wind and solar power? Regardless, if that electricity was used to power vehicles the avoided cost of fuel would be much greater than in the US. I'll let you work the example because I don't know the fuel costs.
Large and constant (and low temperature) aren't absolute requirements, but they rather improve the economics.
I agree that the factor 3 for the cogen/heat pump combination is optimistic.
3 cents per kWh is wholesale. The government mandates an awful lot more than that and there's some other subsidies.
The economics could brighten with a smaller unit like the Honda at the price you cite, because it's so much easier to use a large fraction of that electricity rather than export it to the grid.
Energienetz states that the Honda unit isn't available in Germany yet and being tested.
If you look at the specifications, you'll notice something else, the combined efficiency is 85% for cogen, but efficiency is 95% for heat only operation.
11/17 is 65%, so I presume that the electrical efficiency is 20%.
That depresses the nat gas savings terribly.
Take 100 kWh of nat gas and put it through the co-gen unit, you get 20 kWh of electricity and 65 kWh of heat.
Produce the heat instead via the high efficiency condensing furnace (that comes with the unit anyway for peak loads) and 65 kWh of heat require 68.4 kWh of nat gas.
Generate the 20 kWh of electricity in a combined cycle plant at 60% efficiency and you need 33.3 kWh of nat gas.
Add the two together and you get 101.75 kWh instead of 100 kWh, ie using co-gen instead of separate generation results in a saving of under 2% with the Honda, which is pretty poor.
Post a Comment
I agree that the factor 3 for the cogen/heat pump combination is optimistic.
3 cents per kWh is wholesale. The government mandates an awful lot more than that and there's some other subsidies.
The economics could brighten with a smaller unit like the Honda at the price you cite, because it's so much easier to use a large fraction of that electricity rather than export it to the grid.
Energienetz states that the Honda unit isn't available in Germany yet and being tested.
If you look at the specifications, you'll notice something else, the combined efficiency is 85% for cogen, but efficiency is 95% for heat only operation.
11/17 is 65%, so I presume that the electrical efficiency is 20%.
That depresses the nat gas savings terribly.
Take 100 kWh of nat gas and put it through the co-gen unit, you get 20 kWh of electricity and 65 kWh of heat.
Produce the heat instead via the high efficiency condensing furnace (that comes with the unit anyway for peak loads) and 65 kWh of heat require 68.4 kWh of nat gas.
Generate the 20 kWh of electricity in a combined cycle plant at 60% efficiency and you need 33.3 kWh of nat gas.
Add the two together and you get 101.75 kWh instead of 100 kWh, ie using co-gen instead of separate generation results in a saving of under 2% with the Honda, which is pretty poor.
<< Home
