The Final Word on Steam Boiler Efficiency

[TangentialThreat]

This is a good post with good information that you should take seriously. I found out how inefficient they could be the hard way and in hindsight I should have gone with something smaller.

Caveats:

If you've already gone through the trouble of crafting and heating up a huge HP boiler then you might as well keep it as-is. Invent uses for the extra steam. Melt cobble to magma with the mighty magma crucible, and then centrifuge the lava for infinite metals. Make a forcefield. Put it into energy tesseracts and power the server. Just plan the layout in advance; there are going to be liquiducts coming out of the boiler in every direction by the time you are close to using all the steam.

Is this like the final word on global warming? seems to go on and on for being final!

Le sigh.

 

[Damoklesz]

Combustion engines are really great to use you just need to understand how the Aqueous Accumulator can output it's water, once you master that hooking up Combustion engines are easy.

The AA needs water source blocks on two sides to create water the fastest. The front of the machine counts as a side but you can't hook up pipes to it. That leaves 4 sides left to hook up water piping. If you use Buildcraft Gold Waterproof pipes each side the AA will output enough water to keep a Combustion engine cool. Note that the AA puts pressure to the water so you don't need to pump it out of the block.

Liquiducts have more volume and throughput so you only need to hook up liquiducts to two sides of the AA and that is enough to keep four engines cooled. Remember that the AA puts out water pressure so you don't need a restone signal for liquiducts to output the water.

When I setup Combustion engines I setup everything and baby sit them to max temp and then make sure their water stays topped off and then I don't have to worry about them.

I know I'm getting a bit offtopic here, but people don't seem to mind it in these forums...

Anyways... I've never used waterproof pipes, so I can't comment on that. But with liquiducts, a single connection is enough to pump out all the water an AA produces. This seems to be just the right amount to keep 4 combustion engines cooled. The problems start when you have more engines and multiple AA. In this case the safest solution is to have a separate liquiduct segment for every AA, that way you won't blow up because of the liquiduct limitations. But I don't believe that having 2 connections to a single AA helps you in any way.

 

[Zjarek_S]

Well, when using combustion engines I don't rely on stability of any source of water, if any of my combustion engines reaches orange I shutoff everything.

Back to the topic, finally my 36 LP boiler boiled all fuel from itself and I tested it with biofuel. You can run without problem 36 HP boiler on one bucket of biofuel per 124 seconds, but 36 LP when fed with one bucket of biofuel per 248 seconds will cool down. So high pressure boiler can be more efficient then low pressures ones (eventually ), but I think that advanced solar panels with GT will pay off faster.

 

[Larroke]

The whole idea to boilers is an always on, high efficiency power source. If you don't need the power, then of course it's not worth the investment. Boilers shine when you have the capacity to a build a renewable fuel source! Once you achieve that, the loss of power when you don't need it is irrelevant. Large tanks of reserve fuel and buffer chest/ED-barrel are tied into the system for the sake of if things go wrong (golems get killed, something blows up, world ends, turtles dont' reboot etc) giving you time to notice the issue and repair it. HV solar arrays are the same thing, if you aren't constantly using 512 EU all its power generation is a waste, the advantage of the solar cell (especially with the forestry electrical engine) is that it all fits in a single block with no maintenance... This is also why some see them as cheesy, even with their PITA recipe. A boiler is basically the same thing, but with a great deal larger footprint

If you just want power the moments you need it, then any of the engines stacked sufficiently to your needs will be more "efficient".. its all a matter of preference in the end.

 

[TangentialThreat]

Combustion engines can explode randomly if they are not in the same chunk as their cooling system. The chunk with the engines loads, the chunk with the cooling system does not for some reason, kaboom. We've all probably seen that at least once. You'll learn how to keep that from happening eventually (chunkloaders, building within a chunk), but learning is expensive.

Steam boilers have sufficiently large internal reserves of both fuel and water to survive short interruptions, and are generally idiot proof. When they do explode (it can be done!) the explosion is not actually that big.

 

[WTFFFS]

I refuse to use combustion engines fullstop, they have blown up on me often enough (pre-aqueous days) even when setup correctly that I will not use them again, I'd much prefer to make a couple of dirty great big boilers (with renewable\effectively infinite fuel) than ever deal with those finiky damn combustion engines again. Also given I tend to really really overdo things 10 combustion engines doesn't even come near to running things fast enough for me, my last assembly table setup had 35 lasers and enough power to run them flat out

 

[Hydra]

If you've already gone through the trouble of crafting and heating up a huge HP boiler then you might as well keep it as-is. Invent uses for the extra steam. Melt cobble to magma with the mighty magma crucible, and then centrifuge the lava for infinite metals. Make a forcefield. Put it into energy tesseracts and power the server. Just plan the layout in advance; there are going to be liquiducts coming out of the boiler in every direction by the time you are close to using all the steam.

A single liquiduct can easily transport the steam from a 36 HP boiler to 18 industrial engines. So as long as you plan for the space required for the engines you don't really need a lot of piping at all.

 

[Hydra]

But from what I saw on my laptop install, Two LP's are better than one HP for about a day or two before they even out enough for the efficiency differences to be trivial.

Not "a day or two". 5 hours. 9 minutes. 30 seconds. That is the exact time it takes for a 36 HP boiler to reach full heat. From then on it produces exactly the same power as 2 36 LP boilers and uses exactly the same amount of fuel too. There is no minor difference, it is exactly the same. And that's why I commented on the graphs that the OP showed: they're simply wrong. Efficiency set against time should be identical after exactly 5 hours, 9 minutes and 30 seconds and should not be a curve.

Since you're probably running that boiler on a renewable fuel source the only thing that matters is if you can produce enough to cope with a full heat boiler and if you have enough stocks to have it heat up fully before you run out.

 

[MilConDoin]

And that's why I commented on the graphs that the OP showed: they're simply wrong. Efficiency set against time should be identical after exactly 5 hours, 9 minutes and 30 seconds and should not be a curve.

His graphs aren't wrong, he just uses a different scope. You use the current efficiency at a given point (which stays at the same height after heatup), he talks about average efficiency, which integrates from a cold start to the current point in time, so it will never reach max, but close in to the max-eff line asymptotically.

 

[Omicron]

That's what I told him as well, but apparently he considers any metric that isn't peak temperature fuel consumption not only utterly irrelevant but in fact misleading, for some arcane reason I haven't been able to decipher yet. The argument of "you only use boilers in infinite fuel loops that permanently keeps them at maximum temperature" is not one I'm going to accept, in any case, because it doesn't change anything about the underlying game mechanics or the mathematical results. It merely makes fuel provision trivial. There's no causal relation between the method of fuel provision and the fuel efficiency curve of boilers.

 

[Hydra]

His graphs aren't wrong, he just uses a different scope. You use the current efficiency at a given point (which stays at the same height after heatup), he talks about average efficiency, which integrates from a cold start to the current point in time, so it will never reach max, but close in to the max-eff line asymptotically.

I know, and IMHO that's wrong since it's completely pointless to show the average efficiency over time since after 5 hours why would you even care about it? Don't forget he's kinda overselling this as "the final word on efficiency" and then says that LP boilers are more efficient. No they're not, not in a setup 99.9% of the people have.[DOUBLEPOST=1362586221][/DOUBLEPOST]

That's what I told him as well, but apparently he considers any metric that isn't peak temperature fuel consumption not only utterly irrelevant but in fact misleading, for some arcane reason I haven't been able to decipher yet.

Like I said: the reason is simple: most people only care about the MJ they get per fuel item, assuming that everything has been running for a while. And they want to know if there's a difference between LP and HP. In your whole wall of text that could have been reduced to a single table you didn't answer the question that people actually ask. And since most people who ask these kinds of questions tend to not read whole walls of text they tend to skip to the conclusion and that one is not relevant to their question.

I would not have been so "in your face" if you had not named your thread "the final word on".

Bottomline: you should have made a separation between heat up fuel consumption and 'running' fuel consumption. If you had done that it perhaps could have been considered a "final word".

 

[hotblack desiato]

okay, the answer to that question is a bit buried:

the best setup for a long running boiler is one with 36 boiler blocks. if you don't care about space and iron, go for LP boilers, they heat up faster.

the note regarding those combustion engines: every other boiler than the 36HP is irrelevant, when it comes to fuel as power source.

I'd really like if it is a good idea to use lithium as a fuel source. it's available in large quantities, if the player uses quarries on the redrock biome. else, how about hydrogen is electrolyzing it out of water a good idea? I know it needs a lot of EU, but those could be generated with solars or nuclear reactors...

I know the easiest way is setting up a treefarm or a wheatfarm and process that to biofuel, but honestly? the next step would be something like the aqueous accumulator, but it spits out biofuel and not water.

EDIT: regarding the efficiency: it would be interesting, if there are diagrams which show the efficiency over time for a process like: heat up - run for 10 hours - power down. since the energy from powering down is completely free. and the other part, that was already mentioned in the text: the efficiency-increase is the boiler isn't supplied constantly, but in intervals (heat it up, let it cool down a bit, heat it up again... and so on)

 

[Omicron]

Like I said: the reason is simple: most people only care about the MJ they get per fuel item, assuming that everything has been running for a while. And they want to know if there's a difference between LP and HP. In your whole wall of text that could have been reduced to a single table you didn't answer the question that people actually ask. And since most people who ask these kinds of questions tend to not read whole walls of text they tend to skip to the conclusion and that one is not relevant to their question.

Said table (and more) is made available in the Excel sheet. Which is linked in the conclusion that everyone skips to. I don't see the issue.

 

[baw179]

Can anyone tell me how many buckets of biofuel a HP36 takes to reach max temp please? I know from Zjarek's post above that it will only consume 1 bucketful every 124 secs once up to temp (thanks Zjarek). Unfortunately I have no programs on this machine to access xlsx files (and don't have permission to install anything either) so am unable to view the spreadsheet.

 

[Omicron]

Just shy of 700 buckets, in just over five hours.

 

[brujon]

Omicron:

You're not even taking into account that the steel the HP Boilers take to be produced. It takes 86 steel plates for one 36 HP Boiler, which in turn takes 64,5 Steel Ingots, which effectively equates to 65 because you can't craft in half, meaning you need to burn a stack and a half of coal, turn it into coal coke, and then burn 65 iron into steel, before even beggining construction.

If using charcoal in a blast furnace instead, then you need 4x Charcoal per Steel Ingot, so it equates to 250 extra Charcoal per 36 HP Boiler. Which equates to an extra 416,000 heat units being spent. An LP Boiler only ever uses the fuel needed to burn the iron, if we set that as 1, then the HP boiler, before even producing any amount of energy, already used 3 times as much energy.

And this is energy that will never be recovered, too. Since 2 LP Boilers still take half as much charcoal needed, or twice the amount of coal, depending on whether or not you turned your coal into coal coke first, this is another increase in the permanent gap seen between the two engines in your graph. This is just a thought.

I agree completely with your analysis, the LP boiler is a much more efficient solution for those who don't have some sort of infinite fuel/biofuel setup, and for those who are starting out and still using Charcoal/Peat/Coal and not liquid fuels. However, especially in the context of a server, i have to disagree with you, and say that HP boilers are more efficient.

Space is a precious commodity, setting up 18 industrial's + HP Boiler takes much less space than setting up 2 LP's + 18 Industrial's, especially after taking into account all the tubes you're going to have wiggling around (Unless you want an ugly mess of industrials sticking out of the side of your boilers...)

Both setups accomodate up to 144 MJ/t, enough to power 3 quarries at max power (45MJ/t), and still have about 10 MJ/t to spare. The REASON why it's so popular, is simply because it's the most space-efficient design. Sure, you can stick 24 combustion engines feeding off of a huge iron tank, but you'll also need 8 Aqueous Accumulators for the engines, a whole mess of pipes to connect it all up (which means redstone energy cells, which means more MJ consumption...), and consume about 288 iron just for the Engines, not counting what you need for the 2 LP's! . Or you can have 26 Biogas Engines feeding off of Biomass, get a little bit more in the way of power, need no aqueous accumulators, and produce a little bit more of power, while still keeping it somewhat compact. But then again, you need 286 bronze for that, which means 858 copper, and 286 tin. That's a buttload of copper. In comparison, the LP + Industrial's take in the equivalent of 332 iron. That's 25% of the amount of ores you need to mine. And iron worldgen, as we all know, is now so plentiful that you don't even know what to do with the surplus.

In the end, it also comes down to how you're doing in the way of supplies. Every system has it's drawbacks... And you could argue that the LP Boiler + Industrial's takes in VERY LITTLE in the way of MJ Consumption required to produce the steel plates and iron plates required, that can only be crafted in a rolling machine. So you need to take that into account as well. You can run your Rolling Machine on nearly any amount of power, it only takes 50MJ per operation. 2 Stirling Engine's could run them at 80% efficiency.

So yeah, i have to agree with you on some points, disagree on others, but in any case, it's a very nice analysis, that really debunks some myths. How you choose your power source has to be dictated by the way you play the game, not by what you hear from other players.

FTB has many ways of making resources renewable... One such way is EE3, which, while not as broken as EE2, can be used quite effectively to gather many resources in a renewable manner. Gold or Enderpearls to iron is a popular one, since both can be gotten from Soul Shards for Pigmen or Endermen respectively, and you can even transmute out of WOOD your way into Iron, but that takes a hell of a lot more Minium Stones, which require a lot of mobs...

So, take that in mind, and keep playing =D

 

[Omicron]

Thanks for your analysis, brujon.

Note that I intentionally kept contruction costs out of the analysis. I was trying to model fuel efficiency, not to do a full cost-benefit analysis. Besides, you said it yourself - other ways of generating power also cost quite a bit of resources, sometimes even more in some cases. Therefore factoring in the resource and/or processing costs seems not entirely necessary, because you're going to be spending resources regardless of what method of MJ generation you choose. Also keep in mind that these costs occur only once, while heatup costs occur every time the boiler is heated up. Including the construction cost would skew the performance data for people who already have a boiler built, and are just looking for information. In general I agree, though - building a large boiler and its accompanying infrastructure from scratch can get quite expensive, and should be planned with the same care as its fuel supply.

I'm not convinced that the space argument is that big of an issue... but then again I have never played in the presence of Towny or other such land plot ownership mods. If I did, though, I'd probably just build more vertically. And if DW20 can fit two 36 HP boilers and 36 industrial steam engines in a single chunk on the same y-level with a third of the room to spare, then so can anyone who tries

 

[brujon]

I'd like to add some input over here, on how real world steam efficiency works.

It's VERY complicated, but this graph here http://en.wikipedia.org/wiki/File:Rankine_cycle_with_superheat.jpg has a simplification on how real world steam power generation works.

Basically, right around the 500 Cº mark, you have a sharp drop on efficiency, followed by a large rise in efficiency. Why is that? It's because of the steam saturation. Even when you have steam that's over the boiling point of water, you still get a small amount of water that hasn't reached the gas stage, and this creates drag on the turbines, reducing efficiency. As you rise the temperature, you get a higher saturation point, because more of the steam is gas and not liquid, so you can burn more water, and it gets to a higher speed, with lower drag, spinning the turbines faster. What the graph also shows, is that after the steam has gone past the turbine, it's recollected and turned back into liquid water, so it can be fed again to the machine, so that you won't lose all of your water to the atmosphere.

The last part is not a problem in minecraft, but the first part is. Your analysis shows that CovertJaguar used what's basically a linear progression with power, doubling the amount of possible power production because you generate twice as much steam.

What should happen, however, is that steam supplied from a High Pressure Boiler to any of the Steam Engines should generate more power per steam unit than the amount of power generated by Steam Engines fed steam from a Low Pressure Boiler. On a note, this is part of the reason why Nuclear Power is so efficient in the real world. It can heat up the steam to INSANE levels, creating a huge speed on the turbines, which rotate on a coil generating massive amounts of power.

The problem is, i don't think that's as easy to translate to minecraft, unless you split the steam into two types of steam, low pressure steam, and high pressure steam, and code the engines to output different amounts of power when fed by the different kinds of steam, similarly to how a Biogas Engine generates different amounts of MJ/t when fed creosote, milk, water or biomass.

Another thing you could do, would be to create specific types of piping to Railcraft, that have the ability to detect whether they are connected to a HP Boiler or a LP Boiler, and tell the Engine to output a different amount of power based on what it can detect. The problem with this, is of course, that it would render you unable to use Liquiducts or Buildcraft Pipes to pump steam into your engines, adding another layer of complexity to your build.

So, basically all my rant was about how we need a change in the mechanics based on what you've shown and what i've expanded on, so that the HP Boilers are more than just space savers, and actually become an upgrade to the LP Boilers in whatever size you choose to build them on. What we will see, remains to be seen, but i really do think that the Boilers need some changes in order to make the HP Boiler a clearer upgrade on the LP Boiler, and really, in my opinion, i think the best way would be to make HP and LP steam different "liquids". I believe it's the easiest solution, and the better solution.

I'd also add in that the Industrial Steam engines can only take in HP Steam, and the Hobbyist's Steam Engine can only take in LP Steam, while the Commercial has the option of running on either LP or HP steam, outputting different amounts of MJ/t when fed by different types of Steam. That way, if you just wanted some early power generation, you could go for a 1x1x2 LP Boiler hooked up to a Hobbyist's Steam engine, and later upgrade it to a bigger LP Boiler with Commercial Engines, and when you decided to retire the LP Boiler, reutilize your Commercial's in your HP Boiler, before finally upgrading to Industrial's. So we have a nice technological and industrial Ladder to climb up to.

Since Steam is also used in the locomotive, there could possibly be two different Locomotives taking in LP or HP steam, the first being able to produce it's own LP steam from it's internal engine when fed solid fuel, and the other needing a train station feeding it liquid fuel to produce HP steam and run. It's a nice technological tree that somewhat reflects how it works in the real world you went from locomotives that ran on Coal, to Diesel powered trains, while keeping it in Steam form, before of course upgrading to the Combustion engine (Which is actually less efficient than Steam, TBH, but the reason why we went from steam to straight up combustion is a little more complicated than what we can get on in a minecraft forum)

So yeah, what i'd suggest, is basically what i have already said. One final suggestion i'd maybe add, is make the Steam Engines into multiblocks as well, which i think could actually help them to become more space efficient than it currently is. Bronze -> Iron -> Steel multiblock structures should work rather nicely. I really like this idea, i think it would work out great, making for some really awesome power plant designs, with dedicated power plant facilities and all. Something i'd definitely see the mod being developed into. And when i mean multiblock energy generation, i mean Buildcraft, since EU production is already done in a multiblock fashion.

 

[Omicron]

Unfortunately I don't make Railcraft Such suggestions would be better directed towards CovertJaguar. I can't say I've ever spoken with him, so I can't pass them on for you.

 

[brujon]

I figured, i just wanted to say something lol. Maybe i should log there and comment on the thread.