Industrial Gas Turbine Ignition System

J

Thread Starter

JAMAL

Hi,

I need to know if industrial gas turbine need a continuous spark withing combustion chamber by igniter, or should be enough in commissioning time and after that turbine running automatically?

I mean igniter work each second and make spark in turbine operation.
 
>I mean igniter work each second and make spark in turbine operation.

The English language can be very imprecise. You have identified one instance of it: referring to a gas turbine engine ignitor as a spark plug. Calling a gas turbine engine ignitor a spark plug implies it operates like the spark plug in a gasoline-burning, reciprocating combustion engine, providing an ignition source (an electric spark) on a continuously periodic basis to ignite the gasoline to produce pressure and torque.

In a gas turbine engine the fuel flowing into the combustion area(s) is continuous, not periodic like in a gasoline-burning, reciprocating combustion engine where intake valves, or fuel injectors, control the admission of gasoline into the cylinder periodically. In a gas turbine engine the flow of fuel, once started, is uninterrupted (continuous) and once the fuel is ignited (during the starting process) it is not necessary to have a constant or even periodic ignition source.

It just so happens that a very good ignition source for initially igniting the fuel flowing into the combustion area of a gas turbine engine is a device which produces an electric spark just like a spark plug in a gasoline-burning, reciprocating combustion engine, only physically much larger than the spark plug in a gasoline-burning, reciprocating combustion engine. The ignitor of a gas turbine engine is energized just prior to the admission of fuel into the combustion area and remains energized for a short period of time after flame is detected (to help ensure the flame remains constant at the relatively low fuel flow-rate during the start-up process). But, the gas turbine engine ignitor is de-energized during the start-up process because once flame is well established in the gas turbine engine combustion area it will continue to burn as long as fuel continues to flow into the combustion area.

There are some gas turbine engines that use the igniter to re-establish flame in one area of a staged combustor during load reduction, but the igniter is only briefly energized and is de-energized once flame has been detected in that area of the staged combustor.

The answer to your question is: No. The "spark plug(s)" (ignitor(s)) of a gas turbine engine are not continuously energized whenever fuel is flowing into the combustion area; they are only energized during the starting process to establish flame. But once flame has been detected in the combustion area the ignitor(s) are soon de-energized because fuel flows continuously into the combustion area of a gas turbine engine and, once ignited, burns continuously in the combustion area of a gas turbine engine. In a gasoline-burning, reciprocating combustion engine The flow of fuel into the cylinder of a gasoline-burning, reciprocating combustion engine is not continuous and must be ignited periodically as necessary, and a spark plug is energized when necessary to ignite the fuel in the cylinder, and that is repeated over and over again as the gasoline-burning, reciprocating combustion engine is operating.

To stop a gasoline-burning, reciprocating combustion engine one can EITHER stop energizing the spark plug OR stop the flow of fuel (or both). To stop a gas turbine combustion engine one has to stop the flow of fuel, because once flame has been established using an ignitor (which might be large spark plug) the electric ignition source of a gas turbine engine is no longer required and is de-energized; the flame in the combustion area is the ignition source for incoming fuel. Therefore, shutting off the flow of fuel is the only way to put out the flame in the combustion area of a gas turbine engine.

Lastly, commissioning a gas turbine engine is typically only done during the initial construction of a gas turbine engine and is the process of manually checking each component and system of a gas turbine engine during the construction and initial operation of the gas turbine engine and its auxiliaries and its driven device (such as a generator, or a centrifugal compressor). Commissioning is also often referred to as "start-up"; those activities involved in the initial starting and operation of a gas turbine engine and its auxiliaries and its driven device to ensure the unit operates properly and is properly protected. The subsequent act of starting a properly commissioned gas turbine engine is not typically referred to as commissioning because all those manual activities required to ensure a gas turbine engine is operating properly and is properly protected are only typically done once during the construction and initial operation of the gas turbine engine not each time the unit is started after it is put into service.

Hope this helps. Again, the English language can be very imprecise at times. An ignitor can be a spark plug, and a spark plug is typically an ignition source. But ignition sources for different types of combustion engines, while they might be electric spark plugs, don't all operate continuously or even continuously periodically in all types of combustion engines.
 
G

gustavo_marcelo

Hi.
Appreciate some feedback on below matters.

1. May I know what is the technical reason of GE Frame 9 flame detectors are located at the combustion can 1, 2, 3 and 14. Why does it not equally spaced say at 90 degree each rather than side by side.

Same with the spark plug which is just side by side.

2. Why does the diffusion is said to be a more stable and reliable flame?
 
gustavo_marcelo,

>1. May I know what is the technical reason of GE Frame 9
>flame detectors are located at the combustion can 1, 2, 3
>and 14. Why does it not equally spaced say at 90 degree each
>rather than side by side.

GE has historically (before the advent of DLN combustion systems) usually placed flame detectors in combustion cans near the horizontal joint of the unit (where the various casings are bolted together, horizontally). I believe there was a reason for that placement, and it was because the flame detectors needed to be protected from the heat of the turbine compartment and they could be mounted in open-sided "boxes" placed in the compartment enclosure vertical siding. This also made them accessible from the outside of the turbine compartment. There was usually an expanded metal cover over the open side of the box, which could be easily unbolted and lost. Mounting them elsewhere would have made it more difficult for accessing the flame detectors if necessary and added to the length of the sight tube that connected the flame detectors to the combustion cans. Finally, by mounting the flame detectors on either side of the unit, and not in the same combustion cans as the ignitors (spark plugs) flame could be monitored in cans without ignitors (established through cross-fire tubes). On units with only two flame detectors (many early GE-design Frame 5 units) they were still mounted at the horizontal joint. Lastly, most GE-built units were packaged in enclosures that fit on the same base as the equipment was mounted to. These came to be known as "on-base" enclosures.

It's kind of an occupational hazard of sorts that people new to GE-design heavy duty gas turbines believe every other GE-design heavy duty gas turbine is exactly like the one(s) they are working with. GE has been producing heavy duty gas turbines since the 1950s, en masse, anyway, and there have been LOTS of changes over the years. All GE-design heavy duty gas turbines suck and squeeze air (into the axial compressor and out of the axial compressor), and burn the fuel in the combustors to produce high-temperature compressed air that is then expanded through two or more stages of a turbine, and blow (exhaust to the atmosphere). Most changes were to the auxiliaries and packages (auxiliary equipment and enclosures). And most Frame sizes of GE-design heavy duty gas turbines were very similarly designed and built and packaged--called "scaling-up." It meant, also, that the same control schemes could be used for most all of the designs over decades. Don't fall into that trap--of believing that all GE turbines are "the same," especially when it comes to auxiliaries (and now control systems!).

Then, along came DLN. And, along came cost reductions. And, along came the F-class units, then the H-class machine (short-lived), and now the HA-class machines. And, also, GE bought Nuovo Pignone's and then Alsthom's manufacturing facilities in Italy and France, respectively. And, for tax (avoidance) purposes, mostly, the responsibility for the design and packaging and control schemes and auxiliaries were given over from Schenectady, NY, and Greenville, SC (both in the USA), to the now-GE facilities in Firenze, Italy, and Belfort, France. And a lot of the old, established ways of doing things changed forever. Why? Because they could (be changed).

And, so, you are now finding flame detectors (and spark plugs) mounted in various places around a GE-design heavy duty gas turbine. And one of the reasons for that is that most GE-design heavy duty gas turbines with DLN combustors need more space than was or could be made available in on-base enclosures (because of the additional piping and tubing). So, "off-base" enclosures were designed which had to be erected around the turbine and accessory bases once bolted to the foundation. Off-base enclosures reduced the need to mount the flame detectors near the horizontal joint; they could be mounted anywhere--BUT, because they would be inside the turbine compartment and subjected to the intense heat of the turbine compartment that meant the flame detectors had to be water-cooled. Which led to another whole set of issues, many of which are still being dealt with today. AND, it meant that it was no longer expedient to mount them in the same locations as was done previously (for decades).

There has actually been a move to do away with flame detectors; the same thing can be done with exhaust temperature sensors (thermocouples) and pressure transducers. So, expect more changes, and more changes!

>Same with the spark plug which is just side by side.

Two spark plugs (ignitors) are used simply for redundancy. The placement of them is unimportant, because firing (establishment of flame in combustors where the spark plugs are not located) occurs through the cross-fire tubes. Once flame is established in a combustor with a spark plug, the pressure in the combustor is higher than the combustors on either side, and hot gases flow from the area of higher pressure to the areas of lower pressure and, voila! flame is established in the combustors on either side of the combustor with the spark plug. And, this continues on around the unit as long as there is fuel in the combustors and the cross-fire tubes are working properly (they don't do much, but they can be assembled improperly, and they can become damaged during operation, also). Once flame is established in all the combustors (by propagation, as it's called, through the cross-fire tubes) the pressures in all the combustors are much more equal and there is no further flow through the cross-fire tubes (or at least there shouldn't be!).

Again, two ignitors are used for redundancy only. Only one ignition source is required for establishing flame; two are provided so that if one fails or is weak the other one will provide the spark to get the unit started. (There's just that nagging little worry--is one of the ignitors not working, and if so, which one?!?!?!)

>2. Why does the diffusion is said to be a more stable and
>reliable flame?

DLN flame has a very lean fuel/air mixture, meaning that the ratio of fuel to air is relatively low. This is necessary to reduce the flame temperature to reduce the formation of NOx. There is so much air in a DLN flame that the combustion (flame) is not very stable, especially when transitioning or changing combustion modes. Diffusion flame, on the other hand, has a much higher percentage of fuel versus air, and so it is much more stable at just about any point during the unit's operation.

When changing loads (and combustion modes) in a unit equipped with DLN combustors it's not very easy to change the air flow as fuel flow is changing. So, the ratios of fuel and air can be higher or lower than the most optimal ratio as load and combustion mode changes. The only way to change air flow is by changing the angle of the IGVs, and there is a limit to how much the IGVs can be modulated (opened/closed) during operation (that limit is typically exhaust temperature--the maximum allowable exhaust temperature can't be exceeded during any operating point). While it would be great of every combustor had it's own air flow control mechanism, there is only one way to change air flow to the entire machine (all of the combustors)--and that's using the IGVs. And, because fuel nozzles do NOT always flow exactly the same amount of fuel (even when clean and new!) and because contaminants do get through filters and scrubbers and can cause blockages in fuel nozzle orifices, the amount of fuel in any combustor is not exactly the same as in every other combustor yet they are all receiving the same amount of air flow as controlled by the IGVs.

Changing loads on any gas turbine means changing fuel flow-rates; so with very little chance to change air flow at the same time as fuel flow is changing there are times when the flame is less stable than others (when there is less fuel or more air than can support stable flame (combustion)).

So, DLN flame (called Premix flame in GE-parlance) does not always have the same stability during all loads and operating conditions. And, because the amount of air the axial compressor flows changes throughout the year (as ambient temperatures and conditions change) the stability of Premix flame can also change throughout the course of a year (as ambient conditions change). Throw in inlet air filters (which start out clean and become dirty (choked; plugged) and compressor cleanliness (which greatly affects air flow into and through the axial compressor, and combustors!) and you can run into situations where the flame can be even more unstable than others. It has also become evident over the past twenty years or so (since DLN has been around) that the fuel splits need to be re-tuned on some machine to maintain emissions levels and also to help enhance flame stability.

GE likes to say that DLN technology (especially, DLN-I technology) is a stable and mature technology. That's true, to a certain extent. But, they--and the owners and operators--of DLN machines are constantly learning about DLN technology and its nuances.

So, that's all I've time for today. I'm sure you would like to hear from others instead of me all the time (as would many people, I'm sure). I think most of the control.com GE-design heavy duty gas turbine control readers (and most of the contributors/respondents) are more casual users of the equipment with few years of experience. I have over three decades of experience, and was fortunate enough to come up in a time when people used phones to talk to each other (before, even, email!). We did have TELEXes and then fax machines, which were a huge boon to communications, but being able to talk to people and get to know a little about them, and in some cases, to actually meet them in person and socialize, helped to develop and cement relationships, called "networking" (before the days of the Internet and Ethernet and ARCnet networks) was--and is--invaluable. Having a network of people who could answer questions and provide guidance and even a little of the "background" (such as your Question 1) was invaluable "back in the day." There was still a lot of specialization, but not nearly as much as there is today. (I'm told Siemens does not allow for cross-training of their field service personnel; there are people whose only job is bolting--and they are very good at that, too! But, there are very few people (in the field) that are knowledgeable about the entire Siemens package and how it's supposed to work; they have to continually contact the factory for help with operational issues. Intel, the chip manufacturer, does something very similar. People are assigned very small pieces of chip design or software design and even testing; that way, it's much more difficult for secrets to get out into the world.) So, while there are some very knowledgeable people reading and contributing to control.com, they don't always seem to follow things here as closely as I do.

Lastly, when I started out "in the business," I was extremely fortunate to have had the honour and the pleasure of working with a gentleman (not a GE person, either) who knew just about EVERYTHING there was to know about a GE-design Frame 5 heavy duty gas turbine and auxiliaries, having worked on them in a very remote desert area of the USA for more than two decades. He took me under his wing and mentored me during my first gas turbine start-up--and I learned SO much from that man. He even had little stories like why the flame detectors were mounted at the horizontal joints and other trivia that have proven very handy over the years. In an effort to repay that man for all of his patience and guidance and the knowledge he willingly shared with me I have endeavoured through the years to try to impart some of the wisdom and experience I gained from him as well as the many others in my "network" to as many as possible. That's why I like control.com so much; it's a great way to get information out to a lot of people without having to continually repeat myself (though I do, especially with respect to droop speed control...!).

Excuse me if I contribute "too much." I sort of have an excuse.
 
G

gustavo_marcelo

As usual, you are just wonderful. Wish can meet you in person.

Back to business, if Primary has more fuel and DLN has less fuel, how does it different in term of supplying the demand? I mean is there a different to get 1MW in both combustion modes? If the ratio for Primary is 10:1 (fuel to air) for 1MW ,what will be the ratio for DLN?

Is there any relation between combustion modes and governor modes?

Why does DLN only automatically triggers at certain power level or can it be preselect from inutial starting process?
 
gustavo_marcelo,

Wonderful. But no 'Thumbs Up.' What a fella gotta do?

Terminology and circumstances are both really important in a technical discussion. I think you are confusing "Primary" and diffusion flame. Diffusion flame is characterized by a bright yellow-orange flame--like you see with a match or a candle. For decades, most combustion was accomplished using diffusion flame in just about every type of internal combustion engine (a gas turbine is considered an internal combustion engine) and boilers and water heaters and gas fuel stoves, etc. The problem with diffusion flame is it means the flame is very hot--and produces NOx and CO2 and other undesirable emissions. So, for internal combustion engines (and some other types of devices, too, now) there is Premix combustion which burns fuel at a much lower temperature with approximately the same energy release. So, it takes about the same amount of fuel to produce 1 MW with a gas turbine burning fuel with diffusion flame as with a gas turbine burning fuel with Premix flame.

I also presume we are talking about DLN-I. A DLN-I combustor has two combustion zones: Primary and Secondary. They are separated by a venturi. During starting and initial loading with natural gas fuel, 100% of the the fuel is burned in a diffusion flame in the Primary combustion zone. That is indicated on the Main Display in one of the Status Fields.

When the unit is loaded and gets above a certain combustion reference temperature, TTRF or TTRF1, value (NOT a load (MW) value--a firing temperature reference value), the unit will switch to Lean-Lean combustion mode. In Lean-Lean combustion mode the gas fuel splitter valve moves to approximately 50% position (50% "stroke" in GE lingo), which directs approximately 50% of the fuel to the Primary combustion zone and 50% of the fuel to the Secondary combustion zone. The fuel in the Primary combustion zone continues to burn in diffusion flame, and the fuel in the Secondary combustion burns in both a diffusion flame and Premix flame. Diffusion flame is "visible" to both U-V (Ultra-Violet) and SiC (Silicon-Carbide) flame detectors used on GE-design heavy duty gas turbines.

As the unit is loaded further, and the combustion reference temperature increases above a higher setpoint, the unit will do a Lean-Lean-to-Premix combustion mode transfer. (Some units have a gas fuel transfer valve; some don't, and it affects the exact sequence of the transfer but it's still a Lean-Lean-to-Premix combustion mode transfer.) During the Lean-Lean-to-Premix combustion mode transfer ALL of the gas fuel will be directed to the Secondary combustion zone, most of it burning in Diffusion flame, some of it burning in Premix flame--but there is NO fuel being admitted to the Primary combustion zone, extinguishing the diffusion flame in the Primary combustion zone.

As the Lean-Lean-to-Premix combustion mode transfer continues, some of the gas fuel is then re-directed back to the Primary combustion zone--AND if everything is working correctly, the gas fuel entering the Primary combustion zone WILL NOT re-ignite with a diffusion flame. BUT, it will be burning--just with a lower temperature "flame", called the Premix flame. Premix flame is NOT "visible" to U-V or SiC flame detectors.

When the Lean-Lean-to-Premix combustion mode transfer is complete, approximately 80% of the gas fuel will be going to the Primary combustion zone where it is being burned in a Premix flame. And, approximately 20% of the gas fuel will be flowing to the Secondary combustion zone, where a portion of the fuel will be burned with a diffusion flame (which is detectable by the flame detectors used on GE-design heavy duty gas turbines), and a portion of the fuel will be burned with a Premix flame. This combustion mode is called Premix Steady-State combustion mode.

When the unit is unloaded, as the combustion reference temperature drops below a setpoint, the ignitors (which are located in the Primary combustion zone) are energized to ignite the fuel in the Primary combustion zone into a diffusion flame--and the fuel split (the amount of fuel being directed to the Primary combustion zone) is reduced to approximately 50%, with the other approximately 50% going to the Secondary combustion zone. The unit is again in Lean-Lean combustion mode, with the fuel in the Primary combustion zone burning with a diffusion flame and most of the fuel in the Secondary combustion zone burning in a diffusion flame, with the remainder of the fuel flowing into the Secondary combustion zone burning with a Premix flame.

As the unit is further unloaded, and the combustion reference temperature drops further below another setpoint, the fuel in the Secondary combustion zone is reduced to 0% and the fuel going to the Primary combustion zone is increased to 100%. The unit is again in Primary combustion mode.

When the fuel burning in the Primary combustion zone is burning with a Premix flame it is very lean--meaning the fuel/air ratio is such that there is barely enough fuel to maintain combustion relative to the amount of air in the Primary combustion zone. And, the IGVs are usually pretty far open when the unit is operating in Primary Combustion Mode (unless the unit has and is operating with IBH (Inlet Bleed Heat) in service). As fuel flow (load) is changed in Premix Steady State combustion mode sometimes the IGV angle doesn't change very much, and conversely, sometimes as the IGVs move the fuel flow doesn't change very much. This causes the fuel/air ratio to vary, and because it's already lean that can cause the fuel in the Primary combustion zone to re-ignite with a diffusion flame (resulting in the Extended Lean-Lean combustion mode).

If the gas fuel supply pressure changes significantly during Premix Steady State combustion mode operation, this can result in Primary combustion zone re-ignition (into a diffusion flame). If the IGVs are unstable during Premix Steady State combustion mode operation this can result in Primary combustion zone re-ignition (into a diffusion flame). If the turbine inlet air filters and/or the axial compressor get too dirty, and the ambient air temperature gets high enough this, too can cause Primary combustion zone re-ignition (into a diffusion flame). This is how "unstable" the Premix flame fuel/air ratio is. If the ambient air temperature drops too low and increases the amount of air flowing through the axial compressor, this, too can cause Primary combustion zone re-ignition (into a diffusion flame). If the caloric value of the gas fuel changes, this can also cause combustion problems. And, if the gas fuel splitter valve LVDTs get improperly calibrated, this too can cause combustion problems.

Generally, all things being normal and relatively equal, Premix flame in the Primary combustion zone during Premix Steady State combustion mode operation is pretty stable. But, things can change. And if the unit wasn't tuned properly (tuning is the adjustment of the fuel splits--the amount of fuel being directed to the Primary- and Secondary combustion zones), then it doesn't take much of a change in any of the other variables to cause combustion problems (Primary combustion zone re-ignitions).

There are even Premix blow-outs on some machines--meaning the Premix flame in the Primary combustion zone can be extinguished without the re-establishment of diffusion flame. In this case, raw fuel is flowing through the Primary combustion zone, the venturi of the combustion liner, and into the Secondary combustion zone, where it will burn in a VERY hot diffusion flame--causing serious damage to the combustion liner and the transition piece. If not caught by the turbine control system (usually as an excessive exhaust temperature spread--and resulting trip) very serious damage can occur to the unit.

So, what part of the combustor of a DLN-I combustion system is the DLN area?

To answer your final question, it would be ideal if one could always burn fuel in a Premix flame from starting to Base Load. But, it's not possible. So, we have to live with "staging" the fuel--redirecting fuel to different combustion zones to achieve Premix Steady State combustion mode. DLN 2.6 combustion systems mostly burn the fuel in Premix flame, but not all of it.... The emissions are lower with these systems--but they're not yet available for GE-design B/E class units (which includes the Frame 9E fleet).

Hope this helps!
 
Top