So, I've been thinking about this a lot, and growing peat (WITHOUT using those hideously inefficient farms of Forestry's...) actually seems to be a highly versatile and surprisingly space-efficient method of generating burnable material for energy.
First off, some some useful math for understand just how much yield you can get out of a block of bog earth;
Assuming that bog earth has not changed since the last version of Forestry, that means that 128 (2 stacks) blocks of it are capable of feeding exactly 10 peat-fired engines.
Peat engines consume 1 peat every 4000 ticks, so 10/4000 = 0.0025 peat consumed per tick by 10 peat engines.
This means that 2 stacks of bog earth of capable of producing that much per tick, on average, so 0.0025/128 = 0.00001953125 peat per tick produced per block of bog earth, on average.
That's 0.000390625 peat per second, 0.0234375 per minute, 1.40625 per hour, per block. (assuming you collect the matured earth instantly)
For energy comparisons in boiler heat-value, this is 0.03125 heat value/tick, 0.625 heat/second, 37.5 heat/minute per block of bog earth being farmed
For energy comparisons in generator EU-value, this is 0.078125 EU/tick, 1.5625 EU/second, and 93.75 EU/minute per block of bog earth
I will be using the heat-value per tick for energy comparisons for the rest of this thread, since the HP boiler is my personal preferred power source.
This might not sound like very much, and that's because it's really not. However, that's not what makes bog earth have so much potential. What gives it its potential, is the fact that it has only a single requirement, and that is that it be immediately adjacent to water. No empty space, no lights, not even a block of dirt underneath it, all it needs is water in the immediate vicinity.
This means that you can cram a very large amount of it in an area, up to a maximum potential of 8/9 (or about 88.888889%) space-efficiency. Compare this to other forms of farming;
Sugarcane, which at its most space-efficient, has only a 26.66666667% space-efficiency, and producing a maximum of 1 cane per 1092.32 seconds on average. (The growth values for this is coming from the wiki, by the way) The best way to produce energy from sugar cane, as far as I know, is to turn it into biofuel. Which means you'll be producing 0.1 biomass per cane, (the wiki is actually wrong in its value of .2 per cane, which was changed either in Mindcrack or in recent version of forestry) which is 0.03 biofuel. Multiply this by 36,000, the heat value of biofuel, and you get 1080 heat-value per cane. Divide this by 1093.32 seconds, and you get approximately 0.9887 heat/second, which comes at a cost of about 26.6667% space-efficiency, so that's approximately 0.263659 heat per second, per block of space. Compare this to bog earth in even a 50% space efficient design, (which can be easily surpassed in the automatic farming methods I suggest below) which give 0.3125 heat/second per block.
As for tree and other farms... I really don't know the space efficiency possible with current methods of alternative farming. I know that the old versions of crop harvesters with a turtle can easily surpass almost anything in terms of heat produced per block and time, but that is planned to be removed, so I personally wouldn't consider it. If anyone wants to calculate the space-efficiency of the best forms of farming (even the harvester-turtle ones, since they're currently still valid) please go ahead and post them, and I'll do any additional calculations you're unable to do or are unsure about.
Now, there are still two two problems left in regards to peat farming, one of which is crafting the bog earth in the first place. There are three sensibly sustainable methods for this;
Autocrafting the dirt produced by the bog earth with sand from cobblestone and a water bucket is the simplest and probably the most space and cost-effective way to make it, since you get the bucket back and the recipe creates 6 earth, which will produce enough dirt for of this 1.5 recipe, making it produce a net gain of dirt in total.
Autocrafting the above, but with a wax capsule instead of the bucket is also an option, but requires you to have a sizable bee farm, and also limits your maximum peat production to the size of your bee farm. (which may or may not be a problem for you) This will double the total amount of dirt you have each cycle, so you'll probably want to look into centrifuging it into a small trickle of aluminum and other useful byproducts.
Additionally, the carpenter is capable of producing 8 bog earth with the same recipe as above, but with mulch replacing the water container, and the water instead being taken from its internal tank. If you happen to be over-producing mulch from some other system, such as moistening wheat into mycelium or squeezing apples, then this may be the perfect system for you. Otherwise, auto-crafting with a bucket is more space efficient then making a wheat farm specifically for your peat bog.
The other problem is auto-harvesting it, which I've come up with three methods I will describe below; (Note: is others come up with alternative methods, or wish to post their versions of these methods below, please go ahead and post)
Turtles, which are probably one of the most cost-effective methods, since a single turtle with decently efficient programming can easily maintain around 2k blocks of bog earth, assuming you had a layer of bog earth both above and below the turtle. And given the simplicity of the system, plus the possibilities of resupply stations from Misc Peripherals, you can probably double or triple that amount easily. (assuming you have the horizontal space for it)
Piston-block breaker, which I actually made a thread awhile ago about, which is easily the simplest and most low-tech method to set-up and understand, but also has the lowest space-efficiency of any of these methods. (Not overly much, though)
For info on this method, see the OP of my thread here: http://forum.feed-the-beast.com/threads/alternative-automation-for-wheat-peat-etc.10154/#post-116080
And, finally, the Frame Quarry method. This is the only method I am aware of which allows bog earth to be placed in a nearly-maximum space-efficient manner, but also would probably be the most difficult to make work. I haven't been able to test this idea out in-game yet, but I think this design has the greatest potential for large-scale peat production, simply by the sheer amount of bog earth you can cram into an area. One problem with this method, however, is that it requires you to somehow fit both a deployer and a block breaker in the same space... I do have an idea involving pistons to get around this, but it limits the maximum vertical height of the harvest area to 12 blocks, so if anyone has a good idea of how to do this please describe it below.
EDIT: Wow, that is a lot of words...
First off, some some useful math for understand just how much yield you can get out of a block of bog earth;
Assuming that bog earth has not changed since the last version of Forestry, that means that 128 (2 stacks) blocks of it are capable of feeding exactly 10 peat-fired engines.
Peat engines consume 1 peat every 4000 ticks, so 10/4000 = 0.0025 peat consumed per tick by 10 peat engines.
This means that 2 stacks of bog earth of capable of producing that much per tick, on average, so 0.0025/128 = 0.00001953125 peat per tick produced per block of bog earth, on average.
That's 0.000390625 peat per second, 0.0234375 per minute, 1.40625 per hour, per block. (assuming you collect the matured earth instantly)
For energy comparisons in boiler heat-value, this is 0.03125 heat value/tick, 0.625 heat/second, 37.5 heat/minute per block of bog earth being farmed
For energy comparisons in generator EU-value, this is 0.078125 EU/tick, 1.5625 EU/second, and 93.75 EU/minute per block of bog earth
I will be using the heat-value per tick for energy comparisons for the rest of this thread, since the HP boiler is my personal preferred power source.
This might not sound like very much, and that's because it's really not. However, that's not what makes bog earth have so much potential. What gives it its potential, is the fact that it has only a single requirement, and that is that it be immediately adjacent to water. No empty space, no lights, not even a block of dirt underneath it, all it needs is water in the immediate vicinity.
This means that you can cram a very large amount of it in an area, up to a maximum potential of 8/9 (or about 88.888889%) space-efficiency. Compare this to other forms of farming;
Sugarcane, which at its most space-efficient, has only a 26.66666667% space-efficiency, and producing a maximum of 1 cane per 1092.32 seconds on average. (The growth values for this is coming from the wiki, by the way) The best way to produce energy from sugar cane, as far as I know, is to turn it into biofuel. Which means you'll be producing 0.1 biomass per cane, (the wiki is actually wrong in its value of .2 per cane, which was changed either in Mindcrack or in recent version of forestry) which is 0.03 biofuel. Multiply this by 36,000, the heat value of biofuel, and you get 1080 heat-value per cane. Divide this by 1093.32 seconds, and you get approximately 0.9887 heat/second, which comes at a cost of about 26.6667% space-efficiency, so that's approximately 0.263659 heat per second, per block of space. Compare this to bog earth in even a 50% space efficient design, (which can be easily surpassed in the automatic farming methods I suggest below) which give 0.3125 heat/second per block.
As for tree and other farms... I really don't know the space efficiency possible with current methods of alternative farming. I know that the old versions of crop harvesters with a turtle can easily surpass almost anything in terms of heat produced per block and time, but that is planned to be removed, so I personally wouldn't consider it. If anyone wants to calculate the space-efficiency of the best forms of farming (even the harvester-turtle ones, since they're currently still valid) please go ahead and post them, and I'll do any additional calculations you're unable to do or are unsure about.
Now, there are still two two problems left in regards to peat farming, one of which is crafting the bog earth in the first place. There are three sensibly sustainable methods for this;
Autocrafting the dirt produced by the bog earth with sand from cobblestone and a water bucket is the simplest and probably the most space and cost-effective way to make it, since you get the bucket back and the recipe creates 6 earth, which will produce enough dirt for of this 1.5 recipe, making it produce a net gain of dirt in total.
Autocrafting the above, but with a wax capsule instead of the bucket is also an option, but requires you to have a sizable bee farm, and also limits your maximum peat production to the size of your bee farm. (which may or may not be a problem for you) This will double the total amount of dirt you have each cycle, so you'll probably want to look into centrifuging it into a small trickle of aluminum and other useful byproducts.
Additionally, the carpenter is capable of producing 8 bog earth with the same recipe as above, but with mulch replacing the water container, and the water instead being taken from its internal tank. If you happen to be over-producing mulch from some other system, such as moistening wheat into mycelium or squeezing apples, then this may be the perfect system for you. Otherwise, auto-crafting with a bucket is more space efficient then making a wheat farm specifically for your peat bog.
The other problem is auto-harvesting it, which I've come up with three methods I will describe below; (Note: is others come up with alternative methods, or wish to post their versions of these methods below, please go ahead and post)
Turtles, which are probably one of the most cost-effective methods, since a single turtle with decently efficient programming can easily maintain around 2k blocks of bog earth, assuming you had a layer of bog earth both above and below the turtle. And given the simplicity of the system, plus the possibilities of resupply stations from Misc Peripherals, you can probably double or triple that amount easily. (assuming you have the horizontal space for it)
Piston-block breaker, which I actually made a thread awhile ago about, which is easily the simplest and most low-tech method to set-up and understand, but also has the lowest space-efficiency of any of these methods. (Not overly much, though)
For info on this method, see the OP of my thread here: http://forum.feed-the-beast.com/threads/alternative-automation-for-wheat-peat-etc.10154/#post-116080
And, finally, the Frame Quarry method. This is the only method I am aware of which allows bog earth to be placed in a nearly-maximum space-efficient manner, but also would probably be the most difficult to make work. I haven't been able to test this idea out in-game yet, but I think this design has the greatest potential for large-scale peat production, simply by the sheer amount of bog earth you can cram into an area. One problem with this method, however, is that it requires you to somehow fit both a deployer and a block breaker in the same space... I do have an idea involving pistons to get around this, but it limits the maximum vertical height of the harvest area to 12 blocks, so if anyone has a good idea of how to do this please describe it below.
EDIT: Wow, that is a lot of words...
