Explosion in Frame 7FA Turbine Compartment


Thread Starter

Astrid RamÃrez

Dear colleagues,

We have a gas turbine GE MS7001FA DNL 2.0, 140 MW and we operate with gas or liquid fuel. In this turbine happen the following event:

Operators listened as a leakage. Ten seconds later (approximately), operators listened an explosion in turbine compartment that threw out the first and second level doors.

Immediately, during the visual inspection the operators looked the secondary flexible hose of the combustion chamber # 9 teared from flange of this chamber. In addition, this hose was burned, melted and drilled in a wide section near the side of the secondary manifold. And in the combustion chamber 8, evidence of a burn between the connecting flange and the gasket of the tertiary was found.

The event occurred at the time the unit was working with gas and transferring from Lean-Lean mode to Premix mode (Premix transfer mode), and the unit had a load of 80MW and rising to 90MW. In addition, we are using the atomization compressor instead the purge compressor due to the unavailability of this.

The other secondary hoses were not damaged.

We have photos available for sharing if it is necessary.

Any idea about the possible cause of this event?

Thank you very much.


Is it frame GE 7FA gas turbine because we too have 12 GTS? Can u share me the RCA or root cause of explosion, so that we can have a TBT in the plant.

Are you sure that using the A.A. Compressor as Purge Air is OK? It kind of sounds like you got combustion flow back into the Secondary flexible/manifold. Who said that you could use A.A. instead of Purge Air?. Do have any data from Trip History etc.?

To allow further analysis please post more data and mainly photo data to let us help.
> Are you sure that using the A.A. Compressor as Purge Air is OK? It
> kind of sounds like you got combustion flow back into the Secondary
> flexible/manifold. Who said that you could use A.A. instead of
> Purge Air?. Do have any data from Trip History etc.?

Some further thoughts after looking at P&IDs. If you are running on Gas what changes have you made to the A.A. system to get flow out of the system as it is normally in Bypass mode when running on Gas.

Quote from manual :-
"3. Gas Fuel Operation
For gas turbine operation with gas fuel, the system supplies moderate pressure air for purging the liquid fuel system and cooling the Atomizing Air and Water Injection Nozzles. The system operates as follows under these circumstances: The Atomizing Air Compressor is not running. The air bypasses the AA Compressor via the purge "tee", located between the Moisture Separator and the Air Filter. Before this air reaches the Moisture Separator (PDS1-1), it is cooled to about 225 Deg. F by the water flowing through the co-flow Precooler (HX1-1). Since less air flow is required for the purge mode of operation, the water flow through the Precooler is decreased by the Cooling Water Control Valve Arrangement (VA32-2, VPR68-2, 90AT-1 - See Cooling Water Piping Schematic MLI 0420). This air is at approximately Gas Turbine Compressor Discharge Pressure (CDP) and 225 degrees F. After the air passes the purge "tee', it then passes through fully opened Atomizing Air Bypass Valve (20AA-1), which serves to open and close the bypass air route depending on which fuel is powering the gas turbine. The air then splits into two paths, supplying air to the Liquid Fuel and Water Injection Purge piping and to the Atomizing Air nozzles. The piping to the Liquid Fuel and Water Injection Purge systems pass out of the Liquid Fuel/Atomizing Air Module via connections AA26. The purge air is carried via interconnecting piping to the turbine base (see MLI 0477 Fuel Purge Schematic). The Atomizing Air flow stream exits the Liquid Fuel/Atomizing Air module via connection AA25. Interconnecting piping conducts the air to connection AAI1 at the turbine base. The air then enters the Atomizing Air Manifold, which supplies air to the combustion chambers via 14 flexible hose "pigtails" on 7FA's or 18 flexible hose "pigtails" on 9FA's. The pressure at the connection to the combustion chamber end-cover is approximately 0.985 times Gas Turbine Compressor Discharge Pressure (CDP)."
Another question for you to answer. How long have you been using A.A. for purge? Is this the first Lean-Lean to Premix transfer you have made using A.A. or have you done it successfully before?

If you are running gas, as you are using the atomization compressor instead of the purge compressor, the purge of the atomizing air system provides compressed air for atomization of liquid fuel in the combustion system, also supplies purge air to the liquid fuel and wáter injection Nozzles when the gas turbine is operating on gas fuel, air is supplied to the liquid fuel / atomizing air module from an extraction manifold at the gas turbine compressor discharge.

Hazardous gas protection system is designed so that an explosive atmosphere is not present in the gas turbine compartment under normal operating conditions, if it were a gas leak, the system would have actuated.

An internal explosion can occur when the GCV gas control valves leak, When Gas leak test B does not pass correctly, the tendency is to force the signal to start the unit, when change from D5 (lean to lean) combustion mode to 3 (Premix mode), correct me I am wrong, since I work with dln 2.6, the gas control valve GCV1 (D5) happens to be closed and purge the line, if the valve leak, the gas can be ignited by the purge air From the discharge of the compressor.

Saludos Astrid.
We have been using A.A for purge from february 2016 and we have made Lean-Lean to Premix transfer before at least for five times.

The following is the sequence of alarm and event:<pre>
1 L3GPV EVT Premix Valve Fault Enable
0 L83FXL1 EVT Lean_lean Mode Selected
1 L83FXP1 EVT Premix Transient Mode Selected
0 L3GPV EVT Premix Valve Fault Enable
1 L83FXL1 EVT Lean_lean Mode Selected
0 L83FXP1 EVT Premix Transient Mode Selected
1 L63AT5L SOE Access compartment pressure low- alarm {Q_QD2_CI22}</pre>
For us, the following alarm:
is abnormal, since this alarm is activated at 1 when the turbine starts and reaches 90% of the speed and is set to "0" during unit stop when the speed drops to 90 %.

Thank you
Thank you for the Photos. It surely looks like that you had combustion back into the gas pigtails. I've never seen this happen on a DLN machine but I have seen this happen on a standard dual fuel machine running on gas when the A.A. compressor failed and combustion gases developed back in the A.A. manifold and melted the A.A. pigtails.
I still have a problem with using A.A. instead of purge air and allowing combustion gases to flow back into the gas manifold during transfers. Can you give more information on the use of A.A. and when you started using it?


Dear sir

Can you tell when and which plant so that I can share with management and have a TBT inside the plant for safety reasons.


I don't think we are looking at a gas leak. We are looking at a combustion blow back into a gas pigtail/manifold, possibly caused by a purge valve problem of some kind. Maybe allowing combustion backflow when the purge valves should have been closed.
Any explosion (or combustion) requires an combustible mixture of fuel (and oxygen). It's very difficult (for me) to understand how a single flex hose connected to a manifold which is supplying thirteen other fuel nozzles/combustors at the same time can experience a backflow of hot combustion gases which could cause a flex hose to leak/fail. The pressure of the natural gas in the manifold and all flex hoses should be higher than the pressure of the hot combustion gases. Even if there was no "purge" air flowing in the manifold during the combustion mode transfer, there needed to be a path for hot combustion gases to flow into and out of, which indicates some kind of leak it seems before gas fuel was present in sufficient quantity to be ignited and burn/explode.

It has been mentioned that the problem occurred during a combustion mode transfer, and usually purging is not taking place during a combustion mode transfer (for the systems being switched to/from), so it's still difficult to understand what happened.

Further, if there was back-flow into a single flex hose that would normally result in an exhaust temperature spread (cold spot)--and we don't know what the exhaust temperature spread was during the incident. Was it high, or low, or normal, or quickly changing? What was it before the incident--high, low or normal, or quickly changing?

A big mystery is: What did the combustible gas monitors indicate? It seems that for some reason there was some indication that something was amiss to cause the turbine compartment doors to be opened and a noticeable leaking sound to be detected and yet the unit was not manually tripped. Many of the F-class turbines actually trip on high-high combustible gas detection. What was it that caused personnel to open the turbine compartment doors in the first place? Was it a noticeable leak heard inside the turbine compartment when the doors were closed that prompted the "investigation"?

A leak of combustion gases backflowing into a manifold would most likely present itself as a high exhaust temperature spread--or at least an increasing exhaust temperature spread. What was the exhaust temperature doing at the time of the incident?

Again, there are simply too many things which aren't known about this unfortunate incident. And, using inflammatory wording in the title of the thread without supplying full disclosure and information is misleading and defamatory. The safety record of properly maintained and properly operated F-class gas turbines is very, very good. There have been many documented problems with flex hoses, but, in general with good care during installation and maintenance (including inspection and testing) they are excellent choices versus using hard pipe for connections. How old was the flex hose that failed?

And, why wasn't a detectable leaking sound in the turbine compartment presumed to be the worst possible case (leaking natural gas) and the unit tripped manually and the cause then investigated? Erring on the side of caution is always more acceptable than the alternative. It's not good to trip the unit, but when something is definitely amiss the presumption should be to protect life and equipment not to keep the unit running until the problem makes itself obvious.

A leak in a flex hose could cause hot combustion gases (and air/oxygen) to flow backwards into a flex hose which could lead to an explosive mixture around/outside of the flex hose with sufficient temperature to ignite the mixture at some time if left unattended. But, there would need to be a path for the hot combustion gases to flow to begin with, so some kind of leak in the flex hose before gas was admitted to the manifold/pipe.

But, without a LOT more information and details it's really impossible to say for sure what happened with any degree of certainty. It sure would be nice to have more information, especially about the hazardous gas detectors.... Were they working? What were they indicating? Why was a detectable leaking sound not presumed to be natural gas and left unattended for so long? What was the age and condition of the flex hose which failed? How long since the last maintenance outage?

And, yes, why was AA being used for purge air? Was the packager/OEM aware of AA being used instead of CPD for purge air, and if so, what was their opinion?

I'm glad you joined in.

I know it doesn't sound possible, but if the system went into transfer with purge valves still open, then there is a low pressure path for combustion gases to back flow. Also the fact of using A.A. instead of purge air is suspicious if no changes have been made to the system, in Gas fuel the A.A. manifold can be slightly less than CPD? I've seen the pictures, give me an Email and I'll send them to you. Have a look, it doesn't look like an explosion, more combustion back in the pigtail.