Steam Turbine Tech

[Anonymous]

According to the Classic Trains Turrbines article, the 1938 GE steam turbine had a water-tube semi-flash boiler operating at 1500 PSI and a condensing unit to turn the steam back into water. The boiler is described as flashing water into steam when it encountered a sudden pressure drop(sounds almost like a modern household air conditioner?) The unit was a failure due to complexity and mechanical problems.

We started to get into the problems of water-tube boilers in the compound thread, but would this type of boiler have similar problems or was this whole thing just too far ahead of it's time? Were the steam generators on early diesels similar?

And on the other side of the cycle, why on a traditional steam loco, wouldn't designers condense at least part of the exhaust steam back into water?? Wouldn't running it through tubing in the bottom of the tender be enough to do it?

[jchnhtfd]

Hmm... well, I'll have a go at part of the question, anyway! Some steam engines, notably in South Africa, were built to be condensing. It does save a good bit of water. However... effective condensing requires a good bit more than just some tubing -- you have a lot of heat to get rid of there (technically, more than half of the heat you originally put into the boiler -- but I won't get into why), so you either need a lot of cool water (consider an electrical power plant, or a steam ship) or a lot of air (consider the huge condensing towers at some other electrical power plants). Second, you need to have much better water quality for the boiler than if you don't condense the stuff, as most contaminants tend to build up slowly in condensing units. Third, the condensers are even harder to maintain than boilers! All that being said, though, there were applications (as I noted, South Africa was one) where water was sufficiently hard to get that it was worth the effort.

[tree68]

QUOTE: Originally posted by up829

And on the other side of the cycle, why on a traditional steam loco, wouldn't designers condense at least part of the exhaust steam back into water?? Wouldn't running it through tubing in the bottom of the tender be enough to do it?

If the water in the tender were constantly changing, that might work, but eventually the water would probably approach the boiling point. While that might save energy (because it wouldn't have to heated as much as cold water to make more steam), I suspect it would introduce other problems with things like the injectors. And beyond a certain point, there would not be enough, or any, cooling being done by the water in the tank, so little or no condensing would take place, either.

[daveklepper]

Great Britian's Southern Railroad had the famous Bullard Pacifics which I believe were built originally as condensing engines but were rebuilt after WWII into conventional locomotives . They were good locomotives as rebuilt, and I rode behind one on the "Bournmouth Bell" in 1962.

[Overmod]

Let me toss some points in here.

A flash boiler is also sometimes called a 'once-through' boiler. It is no more dependent on a "pressure drop" to "fla***he water to steam" than any other kind of boiler -- essentially its heating surface can be considered one very long continuous tube, in which sections do feedwater heating, nucleate boiling, and superheating, very quickly. You find these boilers extensively used in steam automobile designs (with the Doble boilers being excellent examples of the principle) and in at least some of the Besler designs for railcars. They are excellent at raising steam quickly, and can be run at substantial pressure with minimal risk of "explosion" due to the small thermal mass of water actually contained in the structure at a given time.

There are other considerations that make flash boilers relatively less useful than other types for steam locomotives, however. The feedwater rate HAS to match the steam mass flow fairly precisely, and the boiler will not work without the feedwater pump in operation -- no injectors need apply. The low thermal mass also translates into limited flexibility of steam generation -- conventional boilers have quite a bit of inherent 'reserve' as the water flashes to steam on even slightly reduced pressure ... this is the effect I think the described quote was supposed to refer to ... which keeps the mass flow high when the throttle is opened without requiring an equivalent change in firing rate. It's very difficult to use solid fuel in a flash boiler -- fairly volatile liquid fuel that doesn't have ashing or sooting characteristics when the flame plume is 'quenched' against some of the tube coils being highly desirable. (E-mail off list for more gory details about this stuff).

Turbines require higher water quality than piston engines -- silica carryover, in particular, being a historic cause of trouble. One of the points of using a high-pressure flash boiler in the UP steam turbines was to have a lower mass flow of steam for the same equivalent shp -- which translated into less mass through the condenser, and that in turn translated into condensers of 'manageable' size and flow characteristics. Note that if you start with very pure water, there are very few places you're going to 'concentrate impurities' by recirculating water through the flash boiler, which has no mud ring, etc. -- the thing requires purified feedwater to begin with (another very real practical problem in steam-era railroading!)

With respect to condensation in the tender water: the question as phrased shows commendable intelligence, because it preserves the closed-loop for purified water in the turbine circuit (the parallel with an open vs. closed feedwater heater would otherwise apply)/ Unfortunately, the absolute amount of condensation you could get this way is severely limited because the latent heat of condensation is so high (and practical condensers need to make the complete phase change back from vapor to water for the *engine* part of the system to work correctly). As a rule of thumb, you heat at least 5 lb. of water to the boiling point to condense 1lb of saturated steam; look at the mass flow through the turbine in lb/hr and then at the amount of onboard water on your engine or in a tender...

Don't think it hasn't been tried historically, either, including the logical variants such as spraying water (which can be relatively impure, btw) on the tubes of an air condenser to increase efficiency! Just too much complexity and not fast enough heat transfer (as it's a factor of surface area and throughput, and the greater the surface area in tube condensers the greater the flow resistance and difficulty with steam/condensate separation, etc. etc. etc.)

[Overmod]

Hey -- steam turbines are Rankine, not Carnot cycle -- with one of the principal differences being that you CAN recover some net cycle energy by doing condensing with the prospective feedwater.

An exhaust steam injector can be thought of as precisely what UP829 was describing in principle: use of the cold tender water as something to condense a fraction of the exhaust steam quickly and effectively. To the extent this is done without impeding the cylinder exhaust, or the feedwater rate (through whatever additional stages of fwh that might be done subsequently) to the boiler, you're getting energy back into the boiler -- useful in every sense -- out of exhaust energy that would otherwise be largely wasted as rejected heat in the exhaust.

I don't mean to make too much of this, as you probably already know it in detail, quite well. But the cycle difference is not just a matter of semantics, and neither is the ability to transfer heat regeneratively as part of cycle economics.

The point about very limited applicability of 'condensing' with finite quantities of available water does remain (and I think I addressed some of it in my previous post on this thread). Something I've played around with is the use of circulating *water* from an open-type steam tender blowdown (via ejector condensers and appropriate pumped hot wells) through air radiators, perhaps under partial or total pressurization (to prevent bubbles or flashing) -- this gets rid of linear dependency of condensation on exhaust steam flow, and gives much better heat transfer to the 'hot side' of the condenser -- water-to-air is tremendously better than steam-to-air, ask any overheated automobile driver...

[Overmod]

OK, here goes:

Point #1 -- REAL steam engines don't exhaust to atmosphere, they exhaust to as good a vacuum as can be arranged. According to my old handbooks, that's about 26" for piston engines, 29" or better for turbines; your numbers will be about the same. If there's any question about the efficiency of vacuum vs. atmosphere, remember that the Parsons turbine on Titanic ran ENTIRELY below atmospheric pressure (courtesy of the excellent vacuum afforded by an effectively infinite-sink condenser, of course) -- and on exhaust steam from multiple-compound piston engines, too -- yet generated a considerable portion of the shp of the ship as a whole.

Condensers that aren't on ships almost always 'impede the exhaust'; the impediment takes the form of perceived back pressure (which can be variable depending on stroke volume, engine speed, etc., and to an extent can be 'tuned' although I don't know how often that was achieved in practice). You may recall the discussion a few years ago regarding the additional back pressure on the UP simple articulateds that used the exhaust steam injector (vs. an open-type fwh). You trade the added thermal efficiency of the latent-heat recapture against the loss of cylinder hp due to the lower pressure drop in the cylinder. Now, this gets a bit interesting because very few steam engines can effectively make use of those last few psi of overpressure in the exhaust steam, perform good 'scavenging' of the residual steam in the cylinder, and not in the process chill the cylinder wall down to where there are problems. Consequently a rather high amount of effective backpressure is not uncommon in simple-expansion locomotives -- whence the "phenomenal efficiency" of the exhaust steam injector aka the 'poor man's feedwater heater'.

To a certain extent, the tradeoff isn't 'apples to apples' -- you're trading thermodynamic efficiency against better performance of the pressure engine -- in other words, against the ability of the cylinders to develop peak horsepower from a given mass flow of steam. You can quickly see that other factors of locomotive operation will rear their heads -- for instance, can the cylinders deliver the 'possible' hp/TE to the rail at a given speed, or could you actually use higher back pressure to limit propensity to slip if otherwise you'd need to keep the cylinder cocks open to get bypass clearance of the admitted steam...

The best 'ordinary' explanation of the difference in the cycles I've ever seen is the one in the Babcock & Wilcox book. The way they introduce the subject is to point out that regeneration adds to the heat drops in the cycle, and then proceed to show all the different places that heat can be transferred according to the laws of thermodynamics and still relieve the heat drop inherent in a simple Carnot-type cycle. I would prefer that they make a bit more discussion out of combustion-gas-side recuperation vs. steam-circuit recuperation (this would, for example, make the difference between an economizer and a locomotive feedwater heater crystal clear!) but any avid student of the technology will start figuring out the practical implications from what they provide.

Something else B&W get to with remarkable, and appropriate, dispatch is the discussion of PRACTICAL thermodynamics (IIRC, they have a clever proprietary term for the subject, like "Croloy" for chrome-containing low alloy steels). They establi***hat it is ECONOMICS that sets the level of heat generation and recuperation that will prevail in a given situation or plant, and give some examples of how to determine the appropriate tradeoffs when there are several different ways to reach an 'optimal' state thermodynamically. Imho, just reading this through a couple of times, and then looking at practical locomotive-driving techniques, would have saved the ACE team a considerable amount of pain and 'dry-hole' engineering speculation (particularly involving practical condenser systems design -- but that's another digression for another day!)

My own opinion on a condenser is similar to the rationale behind some of the French "ORC" (they did it as compounding, but the principle can be applied to bottoming with much greater flexibility) cycles of the 1850s. The object is to remove the heat from the steam with maximal speed, getting the mass flow into the liquid phase as much as possible and as quickly as possible. That takes care of the back-pressure and mass-flow considerations. You then move that heat, in order from highest to lowest temperature flow capability, to all the places it can best be applied ... using materials, heat pipes, pumps, etc. appropriate to the task -- and you do this ASYNCHRONOUSLY to the flow of steam in the engine proper, with reservoirs for stored heat if necessary. While it may seem excessive use of both technology and weight to have a volatile ORC driving the ancillary turbine (in place of exhaust steam pressure like the ACE) you will quickly be able to figure the short-term flexibility and long-term economy that smaller water radiators for tender water can give, vs. large condensers trying to handle mixed steam and water that unpredictably change performance and water demand in even reasonably hot microweather. (There never was a 'tunnel-motor' ACE, and a moment's reflection will tell you there really couldn't be using their stated technological approach, and this quite dramatically crippled the locomotive for even its pathetically inadequate 3000hp nominal output, in the only services for which it would be particularly well-suited...)

[jchnhtfd]

good grief

[dldance]

Just a follow-up on the South African condensing locomotives. That area of their railway was the first to dieselize and as soon as did, they dismantled the condensing equipment from the tenders and ran the engines as conventional steam engines on other parts of their system. Maintenance of the condensers was very expensive - especially since they started with poor water quality to begin with.If you every see photos of late South African steam you may see some engines with very long tenders where the water tank on the tender is much smaller than the loading gauge would permit. These are converted condensing tenders.

dd

[Paul Milenkovic]

Could you condense under part load operations only? Just as you hook up the valve gear to economize on steam when you reach road speed, perhaps one could size the condenser for the part load operation and for starting and hill climbing, require make up water.

[athelney]

QUOTE: Originally posted by daveklepper

Great Britian's Southern Railroad had the famous Bullard Pacifics which I believe were built originally as condensing engines but were rebuilt after WWII into conventional locomotives . They were good locomotives as rebuilt, and I rode behind one on the "Bournmouth Bell" in 1962.

The Bulleid pacifics you saw were not condensing loco''s - they were built with air smooth casing - and a few other gadgets -ie thermic syphons & chain driven valve gear .The only one I can think of that was a turbine was# 6202 of the LMSR called a turbomotive - it was scrapped after being smashed up in the big disaster at Harrow in 1952 -i'think that was the year.

[Anonymous]

If you condense it fast enough, it will create a vacuum.

On naval ships, to add to condenser vacuum, steam eductors are used.

Completely outside of practical consideration, what if you could compress exhaust steam to a liquid state, then remove heat.

Paul's idea of part load condensing is interesting.

What I have read so far about Doble monotube boilers is that though they can go from cold to capacity very quickly, the capacity is limited, and cannot be exceeded. I have also read that while firing rate and feedwater rate have to be closely matched, it is entirely possible.

Modern flat tube technology makes today's radiators far more efficient than radiators of fifty years ago.

Limiting volume of the system makes higher pressures practical.

Higher pressures lowers the latent heat of conversion.

This increases energy density of the system - lowers horsepower to weight ratio.

How much energy into how small a system? How far can these general statements be pushed? I don't know.

These are general statements that I believe that I know enough to confidently make, but to make them more specific, the level of knowledge grows exponentially. But, I'm working on it.

[M636C]

Well, the "Turbine" issue of Classic Trains has reached Australia and I got my copy yesterday.

I was surprised that the failure of the GE Steamotives was put down to complexity and unreliability. In his books on Turbines and on Steam Technology, Wolfgang Stoffels indicated that the GE's were very heavy on fuel, even by steam standards, and didn't meet the economy expectations of the builder. I think he attributed this to scale factors, GE hoping for economy similar to much larger static plants, this being one of the reasons for using the high pressure steam (which the Pennsy showed wasn't necessary for the principle to work).

The use of the locomotives during the war, when economy wouldn't have been as important, but I would have thought reliability would still be necessary, would seem to support this point, that it was fuel consumption, not reliability, that stopped UP 1 and 2.

I'd also question that any Bullied "Merchant Navy" locomotives actually had condensors, although a design based on condenation by mechancal compression of exhaust steam was envisaged and was tested with some success on a 2-6-0 locomotive before WWII. The casing of the Bullied Pacifics would have provided an excellent enclosure for the feed water heating aspects of the condenser system, and might have influenced the styling of the streamlining, but no condenser was ever fitted.

The US Navy tested very high pressure single tube boilers in the two Bronstein class frigates, but the later production ships all had Foster Wheeler conventional water tube boilers.

Peter

[Overmod]

I cannot think of any 'production' condensing express locomotives in Britain. Turbomotive, IIRC, was one of three 'prototype' Stanier Pacifics, and was non-condensing (probably a significant reason for the locomotive's success.

A reasonable page on the Turbomotive is at

http://www.dself.dsl.pipex.com/MUSEUM/LOCOLOCO/turbom/turbom.htm

and Alan Fozard was working on an interesting Turbomotive 2 proposal quite recently.

This locomotive was rebuilt to a more-or-less conventional Stanier Pacific (Princess Anne) when the one-of-a-kind Guy turbine required extensive rebuilding -- and it was in that form that she was involved in the Harrow accident.

Do not forget the "Electro-Turbo-Loco" and its later reincarnations, which were not only turbines but condensing. (Also unsuccessful...)

[Anonymous]

Lots of great answers thanks guys.

On condensers and traditional steam, I was thinking more of extending range by condensing a portion of the exhaust steam, not all of it. Tender size and weight seem to have maxed out on a number of roads at around 20-25,000 gallons. Even recovering 20% would be like increasing a 25,000 gallon tender to 30,000. I suspect NYC-stye track pans wouldn't be practical in open-range country out west. Sounds like the problems aren't worth the actual amount of extra miles one would get.

Regarding flash boilers, turbines, and reserve steam capacity, is this really an issue where the turbine is driving a generator at a relatively constant or a limited range of speeds? The N&W steam turbine also used a watertube boiler years later.

[Overmod]

up829, there are better ways to run track pans than NYC-style flat, including pans which valve the appropriate 'charge' of water into a pan just as the train approaches them. Water has a good amount of inertia (ever watch the progress of sprinkler runoff going down the street?) and hence can be allowed to 'run downhill' in a track pan that isn't level; the relative momentum of the tender scoop will pick it up about as well as if it were calm and flat. Note that it becomes practical (again in a theoretical, not operating sense, so don't start jumping up and down) to heat the water in wintertime, to filter and treat it ahead of time, etc. (since it's not sitting out there in the dusty open between trains)

Part of the reason water capacity 'maxed out' is that water is heavy and if there's an alternative to toting it around, railroads will use it. Another part is that for every 1000 gallons of water you carry, coal (or other fuel) capacity is reduced both volumetrically and by mass -- and you can't scoop fuel from wayside bins or tanks!

The UP exhaust steam ejector can be thought of as a device that gave you some additional water capacity by condensing a portion of the exhaust steam...

The issues with flash boilers et al. DO apply to steam-electrics -- why would you expect this would be any more 'constant or limited range' than seen in diesel-electric counterparts?

Any locomotive using a pressure much above 310psi or so begins to need a watertube boiler -- the Stanley designs, though quite safe, have significant maintenance issues when scaled up to provide the necessary horsepower for locomotives. Remember that a major reason to use higher pressures is that the mass flow of steam (and hence the effective water rate) can be much lower for equivalent drawbar hp, and this also helps 'pay' for the required water treatment, etc. used in the high-pressure systems.

Most of the once-through boilers also involve packaging questions -- smaller size, lower absolute weight, immunity from slosh or weight-transfer issues associated with a large overcritical water mass inside a boiler.


jruppert, you're thinking, and keep it up!

The idea of recompressing the exhaust steam has, in fact, been suggested ... but not for the reason, or to the extent, that you propose. Principal reason is that you'd be inventing something of a perpetual-motion machine ... what would be the point of using steam to compress steam, so you could make more steam? You have to drive pumps to do the compression, and the power to do that comes from where? (Also look at the critical applied pressure to accomplish phase change in steam that is at your chosen gauge pressure without removing latent heat -- it's a Pretty Interesting Pressure!!!)

I used Doble boilers only as an illustration of monotube technology. No locomotive boiler above railcar size would, imho, use this technology. A better solution would be a modified Lamont boiler followed by independent superheater stages.

Speaking of the Lamonts, the design was rejected by the Navy ... but gleefully adopted by the ***. In that service, they apparently were fast steam raisers, ran quite long between overhauls, and provided excellent efficiency.

With regard to the 'maximum efficiency' in a condensing steam cycle, may I suggest small ultrasupercritical engine cycles, where the admitted steam is on the order of 500atm. and is valved via modified fuel injectors. Exhaust can be below atmospheric, into an 'appropriate' condenser. Looking at cycle efficiencies, heat balance, and condenser capabilities and mass flows in such a system can be quite interesting and enlightening.

[Anonymous]

Yes, somewhere well above three thousand psi was what I was thinking, and yeah, I guess that would be perpetual motion, which is why I disclaimered " outside practical consideration". What got me thinking about such an idea was refrigeration cycles.

If conversion of water to steam could be confined to a single small chamber where steam could be directly injected into a cylinder, possibly at an ultrasupercritical state (hint) could you run into the same problem as monotube boilers hitting the proverbial wall of max output?

If such a system were to designed for a high horsepower/load application with large variations in load ( locomotive?), would it make sense to design such a chamber unusually heavy, with very thick walls of excess material with the intent of this excess material providing a kind of "heat battery" to provide a temporary overcapacity to absorb sudden changes in load.

What if this chamber were long and tapered towards its outlet and in the center of an annular heating chamber also long and tapered towards its outlet, and the escaping steam combined with the escaping combustion gases to greatly increase the firing rate of the outer chamber, fuel delivery controlled by pressure drop.

What if several of these where placed on tangents around the circumference of a reaction type turbine?

[M636C]

Overmod,

I think the *** got a bit less than they were expecting with the Lamont boiler in their warships. The "Admiral Hipper" was apparently fitted with this type of boiler, and it is said that its unreliability was the "final straw" that turned Hitler against major surface ships. Now, let me think, was that a bad thing - I'll get back to you when I've thought a bit more about it! The sister ship "Blucher" was sunk in the invasion of Norway, the similar "Lutzow" was sold (incomplete) to the Russians (who said they ever were on the same side) and the Prinz Eugen ended up with the US Navy who left it at Bikini atoll to see what happened. Gives you the impression that the whole idea could be best left alone!

By the way, "Princess Anne" was a really interesting locomotive, a complete "Duchess" chassis mated with the rear extension frames and boiler (only) of the Turbomotive, but using the smaller 6'6" driving wheels (probably only the trailing axle would be original, since the others needed new cranks,either internal or external. I only realised this when I finally saw the drawings of the frames and realized the wheel spacings were different! It may have had the highest tractive effort of any British Pacific.

Peter

[Overmod]

In order:

Refrigeration cycles -- not applicable to water, but much more so to ORC cycles. Tradition has been to rig these so the phase change occurs without the need for explicit compression (which requires more parts and maintenance, forms a critical point of failure for the whole system, and imposes a rate-limiting constraint, in addition to the cycle power loss imposed by that part of the compression heating which is then lost via the condenser). Doesn't mean you CAN'T use a compressor -- just that you have to justify it.

USC engines do not have a wall of max output if properly designed. The question you asked isn't really relevant to USC cycle design: you use a BFP (the technical term is 'boiler feed pump', but the logical alternative acronym cf. BUFF certainly applies!) which in power plants is called a 'feed train' because done in stages for logical reasons) to pressurize the water before heat is applied to it. The 'conversion of water to steam' is then accomplished entirely by heating something with liquid density; the only place the phase transfer will occur is following the injector, in the engine cylinder.

Required 'steam generator' areas for this kind of heat transfer don't have to be very large, which makes their construction relatively 'thinkable' and their operation rather safe. It is of course possible to develop a substantial volumetric reserve of USC "water" (e.g. 7250psi supercritical steam in 'liquid-density' phase) just ahead of the polishing superheater and injector. The max output constraints in these engines are all related to expansion kinetics and limiting speeds in the motors, and practical scavenging and back-pressure in the exhaust.

There is absolutely no need to build a USC motor either heavy or with thermal blanketing (other than to preclude differential heating). There is, however, very good reason to use thermal barrier coatings (e.g. Jet-Hot) throughout the engine and exhaust tract. There is plenty of 'superheat' available in the injection charge, far more than could be transferred from a hot volume of material (at any sane engineering temperature for cylinder-block materials!) to turbulent steam -- and not to the piston and rings that would be sweeping the same transfer area a few moments later...

You do love those tapered chambers! In practice, these engines use very pure water, which is not idly dumped in exhaust. The combustion gas is generated in diffusing burners (enginion's and mine use slightly different design principles) and heat is then recovered via appropriate exchangers in the combustion exhaust, but no cylinder exhaust is ever mixed with the gas. In fact, firing rate of the steam generator and 'polishing' superheater are separate, involving two separate combustion cells and gas flows, in all the practical engine designs I have seen.

You don't control these engines via simple feedback firing circuits. Injector metering demand and a reasonable prediction network for anticipated demand control fuel feed to the burners, cycle time for the burners as they 'flutter', etc., with the result being balanced and recalibrated fairly often. There is no particular need to 'greatly increase the firing rate' even on fairly high net power demand; pressure blowers and draft fans do a fine job of controlling gas speed and temperature, and also help greatly in controlling the various kinds of pollutant that have to be considered (HC and CO vs. NOx generation being one critical flame-temperature tradeoff). It would be possible to use some water or steam injection to control combustion characteristics -- but this wouldn't be the special water used in the actual USC steam circuit...

Using the exhaust from USC cylinders in a reaction turbine is a reasonable idea, as the exhaust flows can resemble the output of a 'puff' governor. As indicated for the Paragon locomotive (see their web site) this could make use of a higher release pressure for the USC engine exhaust than would otherwise prevail for full expansion, and serve much the same use in utilizing steam exhaust heat as a turbocharger turbine does for IC engine exhaust heat. You would almost certainly find the output shaft speed of such a turbine proportionally differing moment-to-moment from the crank speed of the USC motor, however, which would limit the designs and 'missions' for the turbine, probably to ancillary service or electricity generation rather than mechanical propulsion of a vehicle.

If I may suggest: read up a bit on supercritical steam physics, and ultrasupercritical materials research. This will suggest quite a few interesting ideas to you, and stave off some of the less promising ones before you spend much time on them.

[Anonymous]

I'm sorry , I should have been more specific. The Idea that I have with the long tapered chambers is separate from what I was askng about USC cycle engines.

Basically, I had an idea of something along the lines of a gas turbine, where a steam generating chamber would be centered inside what would be the combustion chamber of a gas turbine. I imagined the escaping steam passing through a venturi would provide the airflow in the combustion chamber, eliminating the need for a compressor turbine. The escaping steam and cbustion gases could then act directly on the power turbine. Why not use all of the gas's energy for the turbine instead? I am not thinking in terms of a gas turbine that generates steam, but a steam turbine that can use excess combustion heat in a direct action.

[Overmod]

There might not be much difference in operational thermodynamics between what you're proposing and 'water injection' in some of the old straight turbojets. I don't think you're going to get the 'correct' amount of ram air intake via turbulent mixing with steam to do what you want, though. My guess is also that the heat 'lag' due to having to overcome the latent heat of the feedwater you inject into the chamber will wreck the flexibility of such an engine; you're still going to have to have a 'compressor' to get the ignited fuel burning around your steam-generator elements to get them up to temperature.

Note that you would not need special cans or spacing to operate ultrasupercritical steam-generator elements in the flow of ignited combustion gas from turbine burners (with some care) and this could then be 'bled' to produce impetus and steering jets to direct and cool the combustion gas while implementing momentum transfer in the turbine. Problem is that you retain the need for proportional injectors, and throw away ultrapure feedwater while giving up the better part-load steam consumption characteristics of the external-combustion version of a USC engine.

As far as I'm concerned, a 'steam turbine' using any mix of combustion gas on its blading qualifies as a gas turbine. You could use supercritical water or even superheated steam from the recovery stage of a combined-cycle turbine plant to implement critical flame cooling and water injection in the gas turbine section of the plant ... this basically increasing physical shp and reducing NOx simultaneously. But I don't think using the energy of the steam's ESCAPE (rather than its expansion using heat in the existing turbine gas flow) is a sensible way to augment power, far less replace a compressor.

I will stand corrected if someone provides numbers or formulae to the contrary.

[Anonymous]

Lower emissions could be a usefull benefit from such an application.

It is easy to see without complicated math that you are right about the flexibility of such a system being impaired, and also a compressor would still be necessary to get things started until there was enough heat to generate steam.

Also, when I had suggested designing a USC system with an unusually heavy heating chamber to provide a temporary overcapacity by a heat battery effect, I was thinking of very high temperatures, and understand that would be necessary to yield any benefit. I was thinking of possibly a ceramic material for construction if it could withstand the thermal cycling, if not then the temperature would have to be limited to some subcritical temperature for the given metal of construction.

Anyway, you did say this is not a problem in USC systems.

I guess such dialog can be part of a learning process, but a part of me feels uncomfortable making proposals that are beyond my ability to substantiate. At any rate, this is stuff that I walk around with all day in the back of my head, and I guess it can't hurt to unload.

Just out of curiosity, if a given amount of steam is allowed to expand and do work, then theoreticaly, shouldn't it return to a liquid state at a lower temperature and pressure if compressed, and once compressed could have heat removed more easily? It sounds very Carnot to me.

[Anonymous]

Oh, I just wanted to add that being a diesel mechanic has probably predisposed me to large heavy metalic objects.

[M636C]

jruppert,

Steam condensation by compression was tried on the English Southern Railway N class 2-6-0 no A816 in 1931 to 1935. My understanding is that the principle was tested in a power station, then tried in a locomotive. The theory was apparently successfully exploited, but numerous mechanical problems with the steam driven compressors, and the mechanical draught fans required since steam was not being ejected through a blastpipe, meant that the locomotive was returned to standard condition in 1935. Theoretical work continued looking at fitting a "Merchant Navy" Pacific, but no further practical work was carried out.

This is detailed in "Locomotive Adventure" by H Holcroft, published by Ian Allen Limited (no date, before 1966), pages 155 to 173.

Holcroft was the real designer of the so-called Gresley conjugated valve gear.

I hope this is some help

Peter

[Overmod]

Jruppert, the problem with the 'heat battery' is more a matter of heat TRANSFER speed than of heat INTENSITY (i.e. temperature) per se.

If you're familiar with the characteristics of space-shuttle tile, you'll know that even an enormous temperature differential can lead to very little effective 'heating'.

Likewise, the same anomalously high latent heat of phase change that makes water such a good power transfer agent in an expansion cycle makes it poor for compression phase change. Remember that Carnot is a GAS cycle -- phase changes during the cycle throw off much of the physics. One of the points of Rankine cycle, let me repeat, is in the "recovery" (via thermodynamically-optimized transfer) of much cycle heat that would have to be rejected in a straight Carnot cycle; another is the rather sensible recognition that actual output power, rather than theoretical thermal efficiency, is often the point to be addressed in engine design.

Peter: Holcroft needs no apology, but keep mentioning that he invented the conjugated valve gear, so more people figure it out. (One note on possible credit to Gresley, however; IIRC Holcroft's gear was first applied to 4-cylinder locomotives; I don't have access to his patent (7859, 25. Nov 1909) to confirm this). The choice of a SR Maunsell N class may seem a bit odd -- perhaps less odd when Holcroft's role in their design (which I hadn't recognized until today!) is known...

There was a Duffy paper in the Newcomen Society transactions at the end of the '80s that covered waste-heat recovery on locomotives, and the A816 was mentioned in there.

I also find references that Holcroft covered the subject in the Engineer (v.182, 1946) and in the Journal of the Stephenson Locomotive Society (34, 1958). I have not read either one, and would greatly appreciate if anyone who has these would consider either scanning or copying them -- or selling a copy of the journal outright.

[M636C]

Overmod,

Holcroft's original patent was for three cylinder locomotives, but applied to locomotives with the three cylinders in line, and used two "two to one" rocking arms connected by an oscillating lever which drove the centre valve from its midpoint. The rocking arms were , however, arranged like those on existing Great Western four cylinder locomotives which drove the outside valves from the inside Wascheart valve gear.

Perhaps through language difficulties, it isn't realised that there were far more locomotives with conjugated valve gear in Germany than in Britain and the United States combined, almost by an order of magnitude. The Saxon XVIII class Pacifics had a gear similar to Holcroft's original patent, but the Prussian G12 and G8-3 which were the most numerous had quite a different gear which had been patented by Henschel, and resembled a much stronger version of Gresley's own gear used in GNR 461. However, nearly all the German locos were built before Gresley developed his own gear let alone before he adopted the second Holcroft arrangement.

This came to my notice when the (Australian) Victorian Railways adopted the Henchel arrangement for their 1941 4-8-4 H220, after using a (mirror image) Gresley/Holcroft gear in their 1928 S class Pacifics.

A somewhat outspoken friend of mine (A lawyer, not an engineer) researched the Holcroft and Gresley valve gears, including the original patents, and much of his work appeared in the British RCTS book series on LNER locomotives. He now refers to Gresley as "that lying drunken fool", perhaps not a generally accepted view.

A problem which developed in Australian locomotives with the later Holcroft valve gear arrangement was a build up of ash and grease in the steam passages which greatly reduced their efficiency and power, which might be attributed to the long passages between the angled cylinder and the horizontal valve in the centre cylinder.

Peter