Can 3 GT's Supply 1 HRSG?

[crshears]

Hello all,

I'm new here; this is my first post. If I'm sending it to the wrong forum, please forgive the error and point me in the right direction!

I'm hoping the following may somehow morph into a post-retirement second career; this pursuit is therefore not on behalf of my employer...

I have undertaken a preliminary feasibility study regarding the conversion of a retired steamship [!] with an intact main reciprocating quadruple-expansion engine [but no usable boilers] into a more-or-less permanently-moored combined-cycle powerplant for electrical generation purposes [conceptually fuelled with natural gas]. Dual-fuelling with distillate oil would allow for the purchase of natural gas on an interruptible basis.

The concept I have in mind uses two or three gas turbines [possibly Dresser-Rand KG2-3E's ?] supplying one or more HRSG's [producing saturated steam at around 225 psig] with an aggregate power output of from somewhere between 4 and 6 MW, producing a further 1.5 - 2.5 MW of electricity from the steam engine. The plan in view is to have the vessel serve as a living, working museum with a revenue stream derived from electricity sales that is independent of, and therefore not at the mercy of, the fickle interest of the vacationing public.

I hold a current and valid Certificate of Qualification in the province of Ontario, Canada as an Operating Engineer, Third Class. I have been employed for over thirty-five years [and still am] in the electrical power industry with backgrounds in the operation of fossil-fuelled and water-powered generation, as well as the operation of transmission and sub-transmission systems from 13.8 kV up to 500 kV. It will thus be understandable that I am strong in operational experience but challenged when it comes to the intricacies of plant design.

This is my question:

Would it be feasible to have the two or three GT's supply one common unfired HRSG?

I'm thinking that the gas side of such an HRSG would have to be partitioned into three discrete bays, or cells, or whatever you might wish to call them, so that the thermodynamic behaviour of each section would closely approximate the characteristics of a stand-alone HRSG.

Since the steam engine electrical output would follow the steam production of the HRSG, I'm thinking the steam-to-gas flow ratio of the HRSG would be at least in the general neighbourhood of constant whether one, two or all three GT's were in service. As a consequence I would not expect undesirable pinch or approach point issues, such as steaming in the economizer section, to occur under normal operating conditions.

The underlying reasons for going with one HRSG would be [a] to avoid the unnecessary capital cost caused by the triplication of boiler water level control systems, isolating valves, etc., etc., particularly if a deaerator section is incorporated into the steam cycle, and to simplify control of water chemistry.

Regards, Carl

 

[racookpe1978]

Good question: I've usually seen the reverse. Two (sometimes three) separate CT's + one HRSG/CT feeding either one steam turbine per HRSG, or two HRSG's feeding one ST.

I'm going to let others comment, I don't know this answer.

 

[rmw]

Have you thought about how to isolate individual CT's from the HRSG when down or maintenance is required? How would you handle fluctuating loads? GT's don't like running at partial load, so prudent operation would be to take one or more off line and leave the ramaning one(s) run at higher loads.

with a quadruple expansion engine, how would you reclaim the condensate? HRSG's in CT exhaust streams are going to be a lot less tolerant of oil in the water than the original (probably fire tube) boilers were.

If selling power is your goal, why don't you generate the steam at high pressure/high temperature and send it to a steam turbine generator and extract or exhaust whatever steam is required at 225 psig for the recip?

Just a few thoughts.

rmw

PS: Quadruple expansion - what; is this a sister ship to the Titanic?

 

[crshears]

Hi there RMW, pleased to virtually meet you.

I'll answer your questions and respond to your comments more or less in the order in which you posted them...

You asked: 'Have you thought about how to isolate individual CT's from the HRSG when down or maintenance is required?'

Answer: Yes, I have...

I conceive that each CT outlet would be equipped with what is essentially a very large two-way valve of proprietary design, operable 'on the fly,' to direct the exhaust either into the HRSG or into bypass ducting; the seal faces would be tight-fitting such that this valve could serve as an isolating device. [If necessary, seal air fans could be included to maintain a positive pressure within the valve housing, such that no backflow or gas leakage would occur.]

Incidentally: The capability to operate this valve 'on the fly' would allow for immediate diversion of the CT mass gas flow to bypass in the event of a CT trip, thus avoiding HRSG 'quenching.' The same valve could be used to perform CT start-up, synchronization and loading [including purge] in bypass mode before admitting the CT exhaust into the HRSG, precluding quenching from this source as well.

Assuming three discrete gas paths within the HRSG, a total of six blast gates of conventional design would be placed in the outlets of each CT bypass duct and from the three HRSG gas-side outlets to funnel.

I would place either large plenum 'vent' gates or removable plates immediately downstream [with respect to gas flow] of the two-way valve in both the HRSG inlet and bypass duct inlet to allow for the convective release of heat or [slight] back-flow of leakage flue gas from the two-way valve, if any; these six openings could also serve as access points for maintenance personnel. Six similar vent gates or plates would be fitted at the HRSG and bypass duct outlets, upstream of the outlet blast gates.

I'm visualizing that by providing isolation and venting using the listed devices, a surprising amount of work could be performed without the necessity for a complete plant shutdown.

Your next question was: 'How would you handle fluctuating loads? GT's don't like running at partial load, so prudent operation would be to take one or more off line and leave the ramaning one(s) run at higher loads.'

Answer: The power grid in my province can be considered what is called in electrical terms an 'infinite bus,' since the vast majority of the electrical generation east of the Rocky Mountains in the continent of North America is synchronously connected. Within Ontario, frequency regulation and tie-line control of the system is provided by placing specific entire plants on Automatic Generation Control [AGC] as directed by the Indepenent Electrical System Operator, with the remaining various generators dispatched as needed to meet the load requirements prevailing at any given moment.

As a consequence, I foresee no problem with operating the plant with whichever GT's are in service at the time loaded to between 80% and 100% of maximum, with the output of the steam engine regulated so as to follow the available steam supply. Once running and stabilized in combined cycle mode, the plant would therefore operate more or less at one of three 'pseudo' outputs; either at Step 1 [CT1 + 1/3 SE], Step 2 [CT1 + CT2 + 2/3 SE], or Step 3 [CT1 + CT2 + CT3 + 3/3 SE].

Some of you may already know the following; if you do, feel free to skip over this part...

It is very convenient to think of combined cycle power plants in 'N minus 1' terms. For instance: consider a hypothetical 'three over one' CCPP that consists of three CT's rated at 4 MW and one ST rated at 6 MW, for a total plant output of 18 MW derived from four units. Trying to dispatch the CT's and ST individually poses problems, since their outputs are interdependent. Such plants will instead bid into the market as a plant consisting of three 'pseudo-units' of 6 MW each. There are of course limitations to doing this, mostly due to the time lag involved with raising steam and synchronizing and loading the ST, and whether or not supplementary duct firing is used; but on the whole this works reasonably well.

Now, where was I? Oh yeah...

The generation procurement contract to be negotiated for the facility under consideration would incorporate these power steps. [Other factors would be weekday output versus weekend output, on-peak and off-peak hours for each, five-minute operating reserve availability, black-start capability, the scheduling of maintenance during the spring and fall seasons, which are shoulder demand times for electrical power, etc., etc.]

There would only be part-load concerns if the plant were ever to operate in an electrical island, either alone or with other generation sources; but at this point the operation of what are termed 'non-utility generators' in isolated or island mode is specifically prohibited by the electricity maket rules, unless a 'physical bilateral contract' or other outside-the-box arrangement is put in place.

Next you wanted to know: 'With a quadruple expansion engine, how would you reclaim the condensate? HRSG's in CT exhaust streams are going to be a lot less tolerant of oil in the water than the original (probably fire tube) boilers were.'

Answer: Your statement is quite correct! Exactly how to address that issue is still being investigated. I seem to recall from somewhere that modern materials science has developed ways of enabling engines running on saturated steam to operate without cylinder lubrication...but I may be wrong about that. If I am, I'm hoping that modern materials science has come up with oils that will still properly lubricate cylinder internals but that are much more immiscible than the old-style cylinder oils were, and will be readily detrainable using cyclone separators, semi-permeable membranes, or some other means that will capture the oil for re-use. Final scavenging of oil from the condensate stream would likely have to be performed using some sort of filtering medium. A colleague on the same project has many years of experience operating steam traction engines, locomotives, marine triple-exers and the like, and will almost certainly be able to provide valuable input on this issue.

You also asked: 'If selling power is your goal, why don't you generate the steam at high pressure/high temperature and send it to a steam turbine generator and extract or exhaust whatever steam is required at 225 psig for the recip?'

The answer is, we probably could use a topping cycle; but we would have a tough time fitting all of the necessary gear into the ship without compromising its historical integrity. Plus, combined cycle plants make the best possible use of the energy contained in their fuel [to my understanding only a co-generation CHP plant would be better], and going with a conventional plant and just supplying extraction steam to the recip would likely have a much higher specific fuel cost. Not only that, but there are environmental considerations at play as well; if this is ever actualized, a lot of planning and forethought will have to go in to ensuring not only that the NOx and CO emissions are as low as possible, but that no more than the necessary greenhous gases are produced than needed to accomplish the desired objectives, which are not only to produce power but to have the main engine actually operating for the benefit and pleasure of our guests. In all likelihood a higher ticket price could be charged to see a working century-old powerplant instead of just having the main engine on kinetic display using a turning gear with everything else essentially dead.

One other problem with going your route is that operating engineers of a higher classification would have to be employed, increasing operating costs; depending how you slice it, my number crunching shows that at this point the Chief Engineer for the plant would need only a Third-Class ticket...

You ended with: 'Quadruple expansion - what; is this a sister ship to the Titanic?'

No, she isn't...but she's definitely of the same era. The T's powerplant had a capacity measured to five digits; the horsepower rating of this old girl only uses four.

CR

 

[chicopee]

For what it's worth. In my neighborhood, there is a utility power plant, in Ludlow, MA., with 4 natural gas fueled turbines, each connected to a generator. In turn the product of combustion from these turbines is routed to a saturated steam water tube boiler. This saturated steam is then routed to a saturated steam turbine. Research of this type of installation may assist you.

 

[crshears]

Thanks, Chicopee! Definitely very similar.

Hmmm...was this a plant that was retrofitted with CT's to supply combustion air to an existing conventional watertube boiler? Reading between the lines, it doesn't have the sound of a built-from-scratch CCPP...

CR

 

[MisterDonut]

Two-on-one CTG to HRSG has been done before, and it can work as long as both CTGs are running. If one CTG is down, the "cold" side of the HRSG has reduced circulation and possibly differential expansion. Keep in mind that an HRSG in this size range is going to be a bottom-supported unit, similar to a O-style package boiler; not easily partitioned into seperate gas paths. Also, by the time you construct the elaborate system of diverter valves to make a three-on-one setup work, you might as well have three seperate HRSGs, in my estimation.

 

[racookpe1978]

Sizes matter.

The CT's I usually work on usually range from 115 Meg to 200+ Meg, now getting up towards 300 Meg per turbine. Feels funny to read about a "little" 5 Meg CT. 8<)

 

[crshears]

Howdy racookpe1978,

Whaddaya mean, a 'little' 5 MW CT? I'm talkin' three 2 MW CT's; how's that for tiny? )

I know what you mean,though... a half-dozen of the CCPP's I'm used to dealing with in the course of my work all have CT's rated @ 100+ MW; some are two-over-one, some are three-over-one. There are smaller ones, as well; however even these are in the 50+ MW total plant output range. Were the project I'm investigating to be actualized, I'm pretty certain it would have the questionable distinction of being the smallest CCPP in the province...

While I'm here, I might as well respond to the previous poster, too:

Thanks, MisterDonut! Going with multiple HRSGs after all was where I was starting to go myself, even after initially posing the question...

Out of the gate, I had thought suspended construction could be used, and that solid metal partitions similar to the water-cooled walls in conventional water-tube boilers could be used as gas path dividers. I was also wondering if the downcomers could be placed in the partitions, faired with lagging if necessary to reduce gas-side friction and mimimize the likelihood of steaming within them.

I wasn't anticipating all that much delta T between the various evaporator portions; I had thought that if the inlet and outlet dampers on the out-of-service gas portion of such an HRSG were boxed up, there would be enough cross-circulation on the water/steam side to maintain those parts in a relatively hot state, minimizing delta ex and facilitating rapid loading at high efficiency in the event of a 'scramble' start of any out-of-service CT. In all fairness, though, the same thing could likely be accomplished in the case of separate HRSGs by using auxiliary steam from either a running sister unit or an aux boiler to maintain the temps of the out-of-servce HRGS using a more modern version of what used to be called 'hydrokineters' way back when.

An additional benefit I've thought of since I first posted this is that by using one HRSG the total radiation losses would be reduced, just like having a triplex condo unit that shares dividing walls would have less heating/cooling loss than three fully detached houses...

But everything comes at a price; and what I have realized is that there would have to be one economizer per gas path, even if the evaporator section was common...since attaining proper approach temperatures is dependent upon having gas-to-water-flow ratios maintained within fairly narrow tolerances, even if the evap section piping could be simplified, the economizer sections would still have to be valved in and out of service in lock-step with the CT's...sigh...

Incidentally, since I'm only doing the preliminary study [I'm not qualified to design the plant myself], I'm looking at whether there would be any value in incorporating an electric boiler with integrated accumulator as part of the steam system. One reason I'm considering this is that here in Ontario a smart-meter system has been installed province-wide [except for pockets here and there], and time-of-use metering is used for power billing. Since off-peak power could conceivably be cheaper than the prevailing price for natural gas, and since some manner of standby boiler would likely have to be included anyway, there could be economic benefit to providing steam during off-peak hours electrically.

Another reason is that the plant is conceived as having to operate in two-shift mode, as dictated by the hourly market price for electricity; in order to not 'lose' an hour or two's worth of valuable electricity sales every morning waiting for the engine to be brought back up to operating temperature, there could be merit in performing the engine warm-up [using cheaper ellectrically produced steam] prior to the on-peak period so as to be prepared to promptly generate at full plant output once the on-peak period arrives, maximizing revenues.

I've also heard that modern HRSGs, having relatively small steam drums, have little in the way of steam storage, and that the added flexibility afforded the plant by incorporating an accumulator would enable it to much more comfortably ride through flow excursions during any type of contingency, whether on either the steam supply or demand side.

CR

 

[chicopee]

The power plant in question is in Ludlow,MA which is in central MA..

 

[crshears]

Thanks again, Chicopee!

With some pre-arrangement, I could take a tour over that way on vacation...although I'd have to renew my passport...

CR

 

[racookpe1978]

I wouldn't use electric heaters for "steam" production, but the idea of a electric heater to keep the an off-duty HRSG "hot" while bleeding off steam through the lines to keep the pipes and system hot is attractive. The generator/engine would be on the turning gear, oil be warm, and steam pipes and drains would be "live" .. that's a more manageable load. Remember, steam plants only approach 40 some-odd efficiency. You'd be buying 2.25x electricity units to generate 1.0x electricity units.

 

[crshears]

Hi racookpe1978,

You've captured the concept well. And you're quite right; a snake cannot nourish itself for long by swallowing and digesting its own tail...and now that I've re-read my own post, I realize that I was less than clear: the electric boiler would only be used to maintain the steam plant in a hot condition overnight while not generating electricity.

For further enlightenment, see the thread 'What's Better: VFAC-to-AC & Back or DC-to-AC & Back?' in the Electric Motors, Generators & Controls forum, especially waross' brutally honest assessment of the original idea as posted @ 2119 on 8 April.

All feedback is welcomed and appreciated...

CR

 

[racookpe1978]

Now, iffen I were gonna do a "black pavement" job - not a renovation or add-on to an existing facility - with a 3-to-1 CT to HRSG configuration when different numbers of CT's are going to running irregularly, I'd consider a three-leaf-clover design: three CT's radially aligned to a central single stack with a wrap-around HRSG.

You'd need baffle plates to "swirl" the heat (gasses) from a single running unit into a relatively uniform flow into the central HRSG panels that surround that single central stack. Then, if 2x or 3x CT's are running, the added gasses still "swirl" but are at 2x or 3x the flow rate. All HRSG panels still get hit by the hot gas whether from 1x, 2x, or 3x CT's. No dead air spaces, no too-hot zones always getting blasted by the gasses when the CT's start and shut down. All CT's exhaust evenly so none get a backpressure when other units start up.

 

[crshears]

Intriguing concept...

I'm unfamiliar with the term 'black pavement' job...but the way you're defining it, it seems like this might be one...

The original Scotch marine boilers would be replaced with some sort of CT/HRSG combination, not set alongside them, so neither the term 'add-on' nor the term 'renovation' would apply.

Hmmm...it might still make sense to spool up / shut down the CT's in bypass mode to avoid quenching a hot HRSG...

The only 'add-on' components would be selected electrically driven auxiliary pumps for safe shut-down / redundancy reasons; but inasfar as possible, every effort would be made to use the very well preserved simplex boiler feed pumps, the Weir-type recip wet air condensate extraction pump, the steam-driven circ pump, cross-compound air compressors, steam-turbine-driven fire pump [added in the 1950's following the Noronic disaster], the cargo deck hoisting engine, general service water and ballast pumps, dynamos, etc.

Which company could fabricate a boiler of that one-off design? I must admit it does bring to mind the tangentially-fired [with coal] Combustion Engineering steam generators I cut my teeth on back in my Auxiliary Plant Operator days in the late 70's / early 80's.

CR

 

[MikeHalloran]

Don't bother doing it if you can't get all the necessary parts off someone's shelf.

Even one bit of unique/newdesign kit will consume your budget with unanticipated costs for development/ testing/ replacement with known good stuff cobbled together when it just flat doesn't work.


Mike Halloran
Pembroke Pines, FL, USA