An interesting web page I ran across recently: http://www.bath.ac.uk/~ccsshb/12cyl/ It talks about the Wartsila-Sulzer RTA96-C, a turbocharged two-stroke marine diesel made for large container ships. It comes in six through 14 cylinder versions. Here's some specs for the 14-cylinder model: Total weight: 2300 tons Total displacement: 25,480 liters Maximum power: 108,920 hp @ 102 rpm Maximum torque: 5,608,312 ft-lbs @ 102 rpm BSFC (max power): 0.278 lb/hp/hr BSFC (max economy): 0.260 lb/hp/hr Thermal efficiency is over 50% at maximum economy. Pretty amazing. The page has some great assembly pictures that really get across just how huge these engines are. A complete 14-cyl. engine is 89 feet long and 44 feet tall.49 responses total.
The crankshaft picture is pretty neat - it looks like a standard engine diagram, then I notice the ladder-rungs going down the interior spaces, then I notice the little tiny guys on top...
nice find! I'm forwarding the link to my friends, but how did you find this?
grassolean.com is a harder than usual site to read. try: http://www.montanagreenpower.com/ instead
Re #2: It was posted on a VW diesel owners' Yahoo group I'm subscribed to.
One of Rudolph Diesel's early engines got 75% thermal efficiency (it is claimed in the literature). Sulzer built a single-cylinder, two-stroke, reversible (!), diesel engine with a bore of one meter, which developed 1.47 megawatts, in 1911. (Sulzer was a marine engine developer and manufacturer. There is a history of diesel engine development at http://members.shaw.ca/diesel-duck/library/other/prime_movers.htm, including some pictures of these early big engines
Reversibility is a feature of most two-stroke diesels. It's pretty easy to design any two-stroke engine to run in either direction. Two-stroke marine diesels are often direct-drive and the ship is reversed by running the engine the other direction.
Trivia: The little engines that are used on radio-controlled aircraft are tiny two-stroke diesels.
Oh. Yeah. So they are. That's why the battery is attached to the glowplug and then removed. Cool. :) (I had a line-controlled plane in my youth.)
No, they are carburetted two-stroke engines. The air and fuel are mixed prior to entry into the cylinder. Diesel engines compress only air, and inject the fuel during expansion. Compressed air intake occurs during the latter part of the expansion stroke, when it also scavenges the exhaust out exhaust valves.
Some are, some aren't. Apparently, the one I was thinking of the Cox 0.049 is not a diesel engine. However, there are still a few under 5cc being made. (http://www.iroquois.free-online.co.uk/engines.htm is a list of engines reviewed in magazines. Most of the diesels were reviewed in the 1940s and 1950s.)
How loud can these big engines get? How far away can you feel this thing starting up?
I doubt they are very noisy. At least there is no explosion in the cylinders - the fuel burns as it is injected in a smooth flame. Also, the speed is only ca 150 rpm. You can swing your arm around at that speeds with no pain. I would think most of the noice would come from the compressors, both for the air and for the fuel. I bet, however, they really rumble. Re #10: I think that what they mean by diesel isn't how a Diesel engine is defined. It is the smooth fuel injection during the expansion stroke that is the characteristic of a diesel that was patented. However a carburetted engine, like model engines, can run if the compression stroke causes ignition. This is known as "knock", but doesn't matter at that scale. A glow plug just assists the process with lower compression. There are lots of dictionary definitions of a diesel engine on the web. For example, http://dict.die.net/diesel%20engine/. Do any of those model engines inject the fuel into the cylinder separately from the air, during the expansion stroke?
I don't think so; that'd be too complex. I guess if you want to be picky they're "compression ignition" engines.
I just think we should show some respect for Rudolph Diesel.......
While we're on the topic, here's a two-stroke, radial *aircraft* turbodiesel that looks pretty neat: http://www.zoche.de/ The two-stroke offers some nice smoothness and simplicity advantages over a four-stroke. They're using a pneumatic starting system that sounds pretty interesting, too. It looks like they have 2-, 4-, and 8-cylinder versions planned. Engines like this are rapidly starting to look like the wave of the future for general aviation in Europe, where avgas is hugely expensive compared to jet fuel. There are several companies starting to produce aircraft diesels in various configurations. Besides this one, I've seen pictures of prototypes for a horizontally-opposed 4-cylinder, four-stroke, aircooled engine and an inline 4-cylinder, four-stroke, watercooled engine. The latter is based on an automotive design. Water-cooling is starting to get some attention in aircraft applications again after being abandoned for years. (The standard joke was that using a watercooled engine in an airplane made about as much sense as using an aircooled engine in a submarine.)
It occurred to me that the use of "diesel" for "compression ignition" engines also appears in the term "dieseling" for the continued operation of a carburetted engine if the ignition system fails. I still don't think, however, that Diesel himself ever had anything to do with carburetted "compression ignition" engines.
Still, Diesel should consider himself lucky. No one ever refers to the regular type of spark-ignition engine as an "Otto engine".
And stations would then sell auto otto.
Re #5: I doubt that very much, as the entropy created by combustion alone would limit efficiency to less than that. Re #6: Great page. What the heck do they use to *crank* that thing?
The reported efficiency was *thermal* efficiency, not free energy efficiency.
Re #12: No, "knock" is the ignition of the fuel-air charge from overtemperature before the flame front reaches it. This happens more or less all at once, and causes a shock wave. Knock causes large acoustic waves inside the cylinder, which disturb the stagnation layer near surfaces and transfers a lot more heat to them than they're designed for. Sustained heavy knock tends to be accompanied by things melting or otherwise being destroyed. "Dieselling" in a carbureted car after the ignition is turned off is actually hot-spot ignition. Hot-spots (heated tubes, glow plugs) predate both spark and compression ignition IIRC. This means that model airplane engines share kinship with the earliest, most primitive internal combustion engines. Re #15: Zoche doesn't appear to have anything flying. For a product closer to reality, try http://www.deltahawkengines.com Re #17: Depends if you're in the technical end of the industry or not. Nomenclature matters when you're talking odd cycles; Otto-Atkinson, anyone? Re #0, I'm curious about the lack of modern technical refinements in that engine. For instance, the pistons are oil-cooled. Why cool them, when modern ceramic materials could reduce or eliminate the need for cooling? The heat not lost to the head and piston would help drive expansion, but much of it would come out in the exhaust. This means that there would be a considerable excess of power available from the turbocharger, and that excess could be tapped to push the crankshaft harder. If you can get 50% efficiency from the diesel section, and another 25% from the gas turbine (not unusual IIRC), that's 62.5%. Probably not an insignificant savings if you can get it, so why not? Last, the engine produces 4.27 horsepower per liter. If my SHO did as well, its 3.2 liters would muster 13.7 horses. I'm impressed. ;-) (Actually, I am. At 102 RPM it is producing 1876 joules/liter/rev. Assuming 210 horsepower at 7000 RPM, my SHO only yields 420 J/l/rev.)
Re #20: Rane, you ought to know that entropy created in a cycle has to be removed, and it comes out as unavailability of some kind. Entropy created in the combustion process has to exit the engine as what, if not waste heat? Rudolph Diesel's original engine may have achieved 75% of its theoretical maximum, but 75% net is impossible in a combustion engine as I understand it. If his engine did so well, why are our medium-speed engines getting only about 40% thermal efficiency when our technology is so much better? The best modern equivalent to Diesel's first engine is the cogenerator at the University of Alaska at Fairbanks, which is only betting on 41% efficiency in their coal-fueled engine: http://www.lanl.gov/projects/cctc/factsheets/disel/ccddemo.html After reflection, I notice that the power of this engine is only about 80 megawatts. The single gas-turbine at the Wabash River powerplant in Terre Haute (IN) is 192 megawatts, and is probably a small fraction of the size even when the regenerators are included. Unless the diesel is a lot cheaper for its output or gets substantially better fuel economy than a turbine, I'll bet it's vulnerable. Heck, replacing a 2300 ton engine with a 100 ton engine would allow for 2200 more tons of cargo.
Re #22:
The theoretical maximum efficiency for an Otto cycle engine is
Nth = 1 - r^( 1 - k ).
For any k > 1, including air (k == 1.4), as r increases without bounds,
Nth approaches 1. So an engine can theoretically have any efficiency
that is less than 100%. 75% thermal efficiency would require a compression
ratio of approximately 32 to 1. Normal diesel engines are typically
around 16 to 1, and a bit higher.
A 32:1 engine sounds plausible. It would be necessary to cut in half the
top-dead-center head clearance volume, increase the wall thickness and|or
the material strength by a factor of 2.639 (2^1.4), and also increase
the injector operating pressure by a factor of 2.639. The compression
temperature ratio would also increase, by a factor of 1.392 (2^0.4).
There is, however, little point to it, since this would more than double the
mass of the engine per cylinder displacement, and a 16:1 compression ratio
already affords a theoretical maximum of 67% (89% of the efficiency of the
32:1 engine).
Calm down.
Re #23: The problem with that analysis is that it doesn't reflect the fact that k (the ratio of the constant-pressure specific heat to the constant-volume specific heat, for those still interested) is quite a bit lower in the combustion gases than in the air charge. Gasoline-engine exhaust has k of about 1.27, according to an Allied-Garrett engineer I quizzed once. What this means in practice is that the same expansion of gases does not lower their temperature as much as it heated the air charge, so you're left with lots of heat to discard. As an example, if you assume a constant k of 1.4 (not a good assumption, but not bad for argument) and compress air at 100 kPa and 300 K by 20:1 in volume, you'd get air at 994 K and about 6.66 MPa (megapascals, about 66 atmospheres). Now burn fuel sufficient to heat the gas to 2000 K and reduce k to 1.30. If we assume the same number of molecules of gas, the pressure rises to 13.4 MPa. An expansion of 20:1 reduces the temperature by a factor of 2.46, not the 3.3 times it was heated; temperature falls to 814 K and pressure to 271 kPa. Even a further isentropic expansion to 100 kPa only cuts the temperature to 646 K; that's mighty warm exhaust, and every bit of heat in the exhaust is heat not converted to work. A real engine will have heat losses and friction losses too. That's one of the reasons why you can't get 100% efficiency in a combustion engine, no matter what you do. The 100 MPG carburetor is a myth for solid thermodynamic reasons.
Re #19: Good question. My guess is either compressed air or a smaller "pony" engine, but I'm not sure. I remember finding a page once about marine diesel technology but I can't locate it now. Re #21: My guess is that the reason you don't see ceramics used instead of piston cooling is that the marine engine market is probably pretty conservative. A single engine like this is often used to drive a container ship, with no redundancy. Reliability and long life are the most important design criteria. Ceramics aren't even showing up in car engines yet, as far as I know. In fact, many automotive turbodiesels oil-cool their pistons. Re #22: Fuel cost may be an issue. Marine diesels run on what's varyingly called "marine residual fuel" or "bunker oil". It's a very heavy fuel oil (the gel point is as high as 70F for summer grades) and is very cheap to produce, because it's essentially leftovers from the refining process. Turbine engines usually burn lighter fuels, like kerosine or automotive diesel, which would be more expensive. A turbine design is going to be far more complex, as well -- you need reduction gearboxes (which run at high input speeds, raising reliability issues again), and since turbines can't run backwards you need some kind of reversing gear to drive the ship aft. By comparison, a low-speed diesel can drive a prop directly. Finally, turbines only run efficiently at full output; you can't throttle them back without substantial losses.
I've found a couple references to compressed air being used to start engines like this. Pressurized air is admitted into the cylinders, timed to force the engine to rotate in the proper direction.
For those who are still interested in this, the manufacturer's web site is http://www.wartsila.com. It's pretty dry and doesn't have any of the more interesting (to a geek/gearhead) technical details, but it's there. I'm still looking for details on things like compression ratios, intake and exhaust pressures, inlet temperatures... At least I've found that it does start using compressed air. (30 bar, it says. 1.5 megabytes of PDF to get that.) And the exhaust temperature of their "flex" engine is so low, I don't want to believe it: 285 degrees C. (Where does the excess heat go? Out the cooling jacket?) I'm wondering how well this 50%-efficient engine could be operated on powdered coal, a la the University of Alaska at Fairbanks cogeneration system.
I don't find that EGT too unbelievable, especially if it's measured after the turbocharger. For automotive diesels, a pre-turbo EGT of 1300F is considered the edge of the danger zone, and that corresponds to about 700 degrees C. The turbocharger takes energy out of the exhaust so post-turbo EGTs in automotive engines are usually a couple hundred degrees cooler than pre-turbo temperatures. EGT goes up as you add more fuel, too, so how surprising that 285C figure is depends on the power setting. I'm pretty curious about the compression ratio and how much boost they're running.
It was my impression that auto turbos could run at TITs of 1800 F. Don't the silicon nitride turbines run even hotter? While digging deeper in the brochures from Wartsila, I found that the figures I was looking at were for the flex-fuel engine (fuel oil or natural gas). This engine runs at a fuel-air equivalence ratio of about 0.45, or way, way lean. (They do this for NOx control.) It no longer surprises me that the exhaust is so cool. The thermal efficiency of the 96C series attracts me, though. It gets roughly 50%, and that is without using features like insulated heads and piston crowns. (According to an article on a Caterpillar engine I read some time ago, an adiabatic engine actually loses output from the lower volumetric efficiency of the hot cylinder surfaces, but can make it up due to the greater heat output through the exhaust which is available to run a compounding turbine. U of Alaska-Fairbanks is running a coal-burning diesel cogenerator to heat and power the campus. They expect to run 41% efficiency to start, perhaps 48% once everything is tuned to perfection. Our typical steam powerplants run closer to 30% efficiency, perhaps 33%. If we could replace them with diesels at 50% efficiency, that is 2/3 the coal for the same useful energy. If we can use better technology to boost efficiency to 60% (turbocompounding and a steam boiler to run a bottoming cycle might get that much), that is roughly half the coal for the same output. Half the coal is half the carbon dioxide, among other things. NOx is not a problem for stationary powerplants. It is easily reduced to N2 using a bit of ammonia and a catalyst.
Re #30: A TIT of 1800F is probably possible in gasoline engines for short periods of time. In diesels, though, there's very little afterburning going on so the EGT is pretty directly linked to the combustion chamber temperatures -- so it becomes a bad idea to exceed the melting temperature of any of the engine components. (Some engines use aluminum pistons, so even 1300F would be too high.) Plus high EGTs cause all kinds of wear issues -- it's not uncommon for the exhaust manifolds of turbocharged gas engines to glow cherry red under heavy load, and it's also not uncommon for those manifolds to crack and break up from thermal stresses.
If you've ever seen a dyno stand at an auto company, you'd know that it doesn't take a turbo to have the manifolds and a fair amount of the exhaust plumbing at a glow from dull red to orange. And there's no air injected in the manifolds, so "afterburning" has nothing to do with it; it's just metal that's exposed to the exhaust flow and has neither insulation nor active cooling. The combustion gases in any car engine are far hotter than the melting point of iron, let alone aluminum. But that's neither here nor there. I take it that the potential of this diesel technology for solving other problems doesn't interest you that much? I find it fascinating for stationary powerplants; among other things, it should be possible to start and stop one of these engines in less than a minute, perhaps only seconds. When you compare with the very long ramp up/ramp down times for many steam powerplants, and the reasons why it took Michigan so long to get back on the grid after 8/14, the advantages are obvious to me. One problem with burning coal is the mercury content. I did some research on-line but nothing popped out with the usual mineral form of mercury as present in coal, although I did find a mention that coal washing can reduce it. Of course, reducing the amount of coal used reduces the mercury too.
Almost all the work on mercury from coal concerns its speciation in the gas phase and fly ash. However one would expect HgS to be a common mercury species in coal because of its very low solubility. Another problem with burning coal directly in a diesel engine is solids handling and erosion from abrasive combustion products.
Re #32: It's an interesting idea, but I think gas turbines have higher efficiencies when used for power generation. The throttling and gear-reduction issues aren't as important in a stationary application.
Now I wish that the site I found on the Wabash River powerplant in Terre Haute had mentioned something about mercury removal. It recovers sulfer as H2S and reduces it to the element, but not a word about Hg. The problems with abrasion caused by coal ash appear to be soluble; the Fairbanks powerplant seems to have no show-stoppers. The thing I'd worry about is ash fusion and buildup of slag inside the engine. Oh, another advantage struck me: coal-burning diesels would be the ideal counterpart to large wind farms, because they could be started and throttled very easily as the wind changed. Do you know any decent (free) combustion/engine simulation packages that run on Linux? I've been wanting to try some of these concepts but the lack of a good model always tossed me back to square one (I tried to model an engine in a spreadsheet once, and gave up on the model when it gave me an efficiency of greater than 100% and I couldn't find the bug).
Re #34: Gas turbines fueled by what, though? The Wabash River plant is a coal-fired, integrated-gasifier combined-cycle plant and just barely hits 40% efficiency overall. That's not much better than a typical steam plant at 33%; if you can use a diesel topping cycle to boost that past 50% you've got a big win. If you think about it (and work the thermo), the gas turbine is at a disadvantage because it does not do the combustion in a (nearly) constant volume. The entropy of an ideal gas is proportional to the log of the specific volume, and the gas turbine gets nothing from that expansion in its combustion systems. The diesel can harness the pressure increase and turn more of the heat into work. If I had a decent thermodynamic model I'd try to get the work available from a more-complete-expansion cycle, perhaps with a tuned exhaust system to maintain the turbine inlet at a higher pressure than the turbocompressor. I'd like to have a better idea of the possibilities than I've been able to get thus far, but with the lousy info I've got I can't be sure that my calculations relate usefully to reality.
Minor point - one must oxidize, not reduce, sulfur as H2S, to obtain elemental sulfur.
I knew that mercuric sulfide was not very soluble (I found it was used as a tatoo pigment), but I can't believe I mis-spelled "sulfur". I've tried finding the reaction used to convert H2S to pure sulfur, without success; I wonder if it yields anything useful, like H2. No suggestions for engine simulations?
The H2S is oxidized partially to SO2, which then reacts with the remaining
H2S by the disproportionation
2H2S + SO2 = 2H2O + 3S
The oxidation is catalyzed and there are several difrerent implementations
of this process.
So it starts with the reaction
H2S + O2 = H2 + SO2
right? So some free hydrogen is produced?
No free hydrogen is produced. The oxidation of H2S produces only H2O and SO2. In fact, I do not know of any oxidation reaction that produces free hydrogen so long as free oxygen is present. Water is thermodynamically very stable and is strongly preferred as an oxidation product.
Yeah, I remembered that later. It takes a little bit of heat to start the hydrogen-oxygen reaction, but the reaction is exothermic, so it provides the heat necessary to keep itself going.
There are bacteria which get their energy from the oxidation of hydrogen sulfide (some of them produce large amounts of solid sulfur as a metabolite). It seems possible that some mechanism could recover useful energy from the combustion of H2S. Hmmm.... some fuel cells run at well above the melting temperature of sulfur... could be interesting if it wouldn't poison the catalysts. All you'd need is a sufficiently reducing environment on the fuel side and you'd recover molten sulfur as the byproduct.
There isn't much H2S around to use as a fuel. It is also more toxic than HCN. I don't think it will be an optimal fuel. The more common bacterial use of sulfur compounds is of SO4(-2, sulfate), which they use as an oxidant, reducing the sulfur from +6 to -2 (as H2S). This is the source of H2S in much of deeper Michigan groundwater.
If you are gasifying coal or "sweetening" natural gas, there can be plenty of H2S around. Certainly enough to be a problem. Disposing of solid sulfur is probably easier and cheaper than other possibilities like gypsum. The Wabash River powerplant page claims 30,000-odd tons of sulfur recovered during its test period. If we assume 10,000 tons a year of sulfur taken from H2S, the yield of hydrogen would be about 625 tons a year or 1.7 tons of H2 per day. At 61,000 BTU per pound that's over 60,000 KWH per day, or about 2.5 megawatts continuous. Gas wells probably handle a lot less H2S, but a few tens of kilowatts might still be welcome; production of a valuable byproduct such as sulfur rather than a waste product like gypsum might be worth it too.
Have you said how you are going to get H2 from H2S? Yes, sulfur from desulfurization is supplying most of our needs for sulfur - much more than from Frasch plants. But you cannot yet economically recover H2 from H2S, although there is ungoing research (http://members.tripod.com/sulfotech/h2scrack.html). The main use for all that sulfur is for producing sulfuric acid. I don't think there is any surplus that needs to be thrown away.
I wouldn't get H2 from H2S, I'd convert H2S to H2O+S
via an anode reaction like this:
H2S + 2 OH- -> 2 H2O + S + 2 e-
As long as there was H2S present, the environment would be
too reducing to allow SO2 to form.
This depends on a catalyst which isn't poisoned by sulfur,
of course, but given our progress in genetic analysis of
extremophile organisms (and the tendency of those organisms
to live on things like H2S) I'm not sure that this is going
to be a huge obstacle. You might be able to find a lot of
the enzymes in bacteria growing at hydrothermal vents.
In #45 you said "the yield of hydrogen would be about 625 tons a year or 1.7 tons of H2 per day". That sounded like getting H2 from H2S. What did you mean, then? "As long as there was H2S present, the environment would be too reducing to allow SO2 to form" means that SO2 is formed, but reacts immediately with the excess H2S. This is the process I described earlier (I now recall it is a form of the Claus process, which is conducted in the liquid phase and does require a catalyst to be practical. No electrolysis is required.
Rane, if you look at the half-reaction in #47 you'll see that H2 need never be formed. The energy required to dissociate H2S into S and bound hydrogen (which later combines with hydroxyl) would reduce the voltage somewhat, but the "fuel" is free. I have no doubts that this would yield energy, because there are already aerobic bacteria which metabolize H2S.
You have several choices: