Dear all

During startup of 9FA machine on fuel oil, unit tripped due to false start drain high temp (upstream VA17-1a) so valve checked and it is NOT passing and this happening only with fuel oil.

- Temp was increasing gradually not suddenly from 95 to 117 C
- After starting water injection the temp increased 5 degrees

So what are the possible reasons?

 

Kamel,

Was there a failure to ignite (or several) prior to the fired start?

Without being able to see the P&IDs for the False Start Drain system for your site, we can't know which particular drain valve detected the high temperature.

Are you saying that a single T/C or RTD (or transmitter using an RTD or T/C for sensing) indicated a high temperature and initiated a turbine trip? That kind of goes against the concept of using redundant sensors to protect against false trips, but, hey, with F-class machines all kinds of "standards" have been re-standardized. Especially on units that come from or are engineered in Belfort.

Not being able to see how--and where--the sensor is mounted on the False Start Drain line (before or after the valve; inside or outside the turbine compartment), have you considered that something is abnormally heating the sensor during liquid fuel firing? Some vent fan that's not running, or some gravity-operated damper nearby that's stuck closed or has just started not working properly?

Again, without being able to see where the sensor is located, is it possible that somehow water or fuel oil is getting down into the line and getting heated as the unit is accelerating and loading? You didn't say where in the "start-up" the trip occurred, but it was some time after generator breaker closure because you mentioned water injection. That could explain why the temperature is increasing rather slowly. Even a small flow of compressor discharge air or combustion gases through the valve (which should be closed at some point during the start-up--but is open for some time during the start-up also!) would take a while for the line--and the sensor--to heat up. Again, if there was one or more failed liquid fuel start attempts prior to the fired start, there may have been some "residual" fuel that made its way down the line and got heated during starting/loading.

You say the valve is NOT passing. Was it physically removed from the line and tested with water under pressure to prove it's not passing?

How long since the last maintenance outage when people were climbing around the unit, using piping--especially electrical conduits--as ladders and scaffolding? Is it possible the sensor was damaged?

Again, I would really like to know if a single sensor can initiate a turbine trip. I know hot combustion gases leaking through a False Start Drain valve have caused at least one fire, so it's a critical parameter to monitor. But, a single sensor used for tripping?

Use your P&IDs to understand where the False Start Drain valve line that sensed the high temperature is draining. The combustion wrapper? The lower combustion cans? The turbine shell? Then try to determine how some liquid might have gotten into the line and gotten heated during starting.

Another question would be: What were the exhaust temperature spreads when the trip occurred? If there was a high spread, then it might be that either a fuel nozzle is not working properly and fuel isn't getting atomized properly and is "dribbling" liquid fuel into the area and it's making its way to the drain line/valve. Or, that some water injection nozzles in one or more combustors are plugged and not atomizing the water properly and they are dribbling water into the area which is making its way into the line/valve.

Or, just maybe, the False Start Drain valve in question did NOT fully close on that start attempt--but it did when you tested it--and a small amount of hot air/combustion gas was being passed and the condition was real.

 

Each can of the lower 9 canes has one drain line without valves and the 9 drain lines connected together and 3 RTDs mounted on this common pipe before the False Start Drain valve in question .so RTDs are in between the valve and cans common drain line . Distance between valve and common line around 90 cm (inside compartment)
Valve was tested for leak at 80 MW (there is a downstream 3 way valve converting the flow to one drain pit)

- No failure to ignite prior to the fired start happened

- There are 3 RTDs working properly (as per I&C)

- Before and after this fuel oil trials gas fuel used without any problem so if there is any problem due to fan failure or RTD damage or any leak it will affect in both gas and oil fuel

- Unit tripped at around 100 MW

- If there is some residual water or oil fuel in pipes how it will heat up , if heat will transfer by conduction from comp discharge so in normal operation the air trapped in this pipes will be heated also (thermal capacity for water is higher than air)

- exhaust temperature spreads when the trip occurred were normal

 

Kamel,

With respect to the Frame 7 units, in addition to the SOVs there are three manual drain valves which are ported to the False Start Drain Tanks. Two of these are located under the turbine casing and if inadvertently opened, would port hot gas into the drain tank. We have had this happen in the past and have replaced the valves with a locking style valve to prevent the valves from be unintentionally opened. I'm not sure if this is the case with the Frame 9s but thought it may be worth looking at. Let us know what you find.

Mark Emmerth
Duke Power

 

kamel,

Ah, so good to know that redundant RTDs were used to detect the temperature--which means there was, in fact, a "high" temperature in the piping. I would be grateful to know how the sensors are attached to the pipe or inserted into the pipe (in thermowells?).

My best guess based on the information provided is that <i>either</i> there was some liquid (fuel; water) trapped in the piping and it was heated by the temperature of the compartment environs, <i>OR</i> the valve did not fully close during the start attempt. Most of the piping used for those systems is plain carbon steel piping, and flakes of rust can prevent the valve from closing fully. I'm sure the valve opened during the trip and re-start, and then closed again during the subsequent start when it was re-tested.

If you have some kind of historian, can you see when the drain line temperature started increasing during the start?

If there was any liquid in the combustion cans or piping (from fuel that had not burned during the initial firing) it would be heated by compressor discharge air. Also, unless the drain lines from the combustion cans to the common line are lagged (insulated) the ambient temperature in the area (which can be quite hot when the turbine is running) can heat the piping which would heat the liquid inside that was prevented from draining out by a closed False Start Drain valve.

I also believe that the combustion cans are hotter during liquid fuel firing because there is no DLN premix for liquid fuel--that's why water injection is required for NOx emissions reduction. In DLN mode on gas fuel, the hot gas temperatures are MUCH lower--which is what prevents the formation of NOx. This is because the air/fuel mixture is extremely lean.

When liquid fuel is being burned--even in a DLN combustor--there is no premix mode, no extremely lean air/fuel mixture. (It's leaner than normal, but not like when gas fuel is flowing.) So, the flame temperature is higher (because all of the fuel burns with a diffusion flame) which means the NOx is higher than when gas fuel is burning in premix mode, which is why water injection is required. Water doesn't reduce the flame temperatures by that much, but it does reduce it sufficiently to reduce NOx formation (which increases with flame temperature).

So, without being able to see your installation and any historical data, it's likely we can't be of much more help.

It's not a control system problem though. As you said, there wasn't any problem when running on gas fuel. It may be that the temperature setpoints for the leak detection are set a little low, but that's a matter to be taken up with the OEM/packager.