Hi all

I heard today something about flame detectors in mark5 and older GE speedtronic control systems (mark 1_4). In older speedtronic they are just switches and if fire established they act (with any power of flame). but in mark5 we need a minimum level of flame so they can act?

Can any one guide me more?

 

dontknowman,

Didn't anyone ever tell you you shouldn't believe everything you hear?

The most common type of flame "sensor" (detector) used in early Speedtronic turbine control systems, from Mark I through Mark IV and into the early production of the Mark V were Geiger-Muller-type Ultraviolet flame detectors (most were manufactured by the Honeywell corporation). In this type of sensor UV flame intensity (such as is emitted by diffusion flame of combustion of fossil fuels) causes the output of the sensor to vary--more UV means a higher output frequency; lower UV means a lower output frequency.

Early gas turbine control systems were only concerned with the presence of flame, and not with the intensity of the flame. So, even though the flame detectors could provide information on flame intensity all that was necessary was to know if flame was present or not. So, when the output of the sensor reached a certain minimum value then a discrete (logic) signal was generated to indicate the presence of flame, and when the diffusion flame UV intensity dropped below a minimum value then the discrete signal changed state to indicate no flame (loss of flame). Even though the sensor output might vary with flame intensity as unit load (fuel flow-rate and air flow-rate) changed, once it was above a very low level and the logic indicating flame was set it never changed until the intensity dropped to a very low level (almost zero).

For example, the Mark IV turbine control system used a third-party interface module to provide the voltage required by the flame detector (approx. 335 VDC) and to produce a discrete (logic) output to the control system to indicate the presence or absence of flame.

When the Mark V was being developed DLN combustion systems were in their infancy and it was thought that flame intensity would be a good thing to monitor. So, instead of using a third-party interface module between the flame detector and the control system the Mark V was designed to provide the necessary voltage for the flame detector (still approximately 335 VDC) and to be able to monitor the output of the flame detector to detect diffusion flame UV intensity as well as to produce a simple logic to indicate the absence or presence of flame.

Later in the production of the Mark V, GE acquired Reuter-Stokes and their Silicon-Carbide line of flame detectors (called 'Flame Trakkers). These devices only required a two-wire 4-20 mA (nominal 24 VDC circuit) to operate--but the Mark V could ONLY read the frequency output of the Geiger-Muller-style flame detectors. So, R-S had to invent an interface module to produce a variable frequency output from a 4-20 mA input to supply to the Mark V's flame detector inputs. The Mark V still had internal "switches" (comparators) to detect the presence or absence of flame from the R-S module output signal.

The design of the Mark V required flame detector inputs to be connected to the PTBA on the <P> core--and they couldn't be connected anywhere else. And, the level of intensity required by the Mark V was virtually identical to that required by the third-party Geiger-Muller UV flame detector interface modules.

The Mark VI and the Mark VIe were designed such that the flame detector inputs were not limited to any specific input circuit/card. So, the R-S 4-20 mA Flame Trakkers could be directly connected to any 4-20 mA input and used for flame detection--and flame intensity monitoring. And, for retrofit applications, the Mark VI and Mark VIe can accommodate Geiger-Muller UV flame detectors (and produce flame intensity indication from them in addition to the ON/OFF indication).

In my personal experience I've never seen flame intensity provide any meaningful troubleshooting data--none that was more informative than just the presence or absence of flame. In fact, flame intensity indication caused more questions and grief and heated discussions than I care to remember. Operators, and their supervisors, seem to believe that all combustors should have exactly the same flame intensity and any intensity which is not the same as every other intensity indicates a problem with the turbine control system which must be immediately resolved as it will cause damage to the turbine. Which is so far from the truth it's not even funny. Unless every orifice in every fuel nozzle is EXACTLY the same size (and none ever are), and the amount of air entering every combustor is exactly the same (and it never is) AND the flame detector is perfectly aimed at the center of the flame ball (and it NEVER is for all loads and operating conditions) AND every flame detector is identically manufactured and has exactly the same tolerance (and they NEVER do) then no two (or four) combustors will indicate the same intensity. It's just NOT going to happen.

Even when high exhaust temperature spreads exist, it's very rare that the combustor(s) experiencing the problem(s) have flame detectors mounted on them, and even if they do I've never seen an intensity indication that proved beyond a shadow of a doubt that a particular combustor was experiencing a problem (unless there's NO flame in a combustor with a flame detector).

I've had to provide flame intensity readings to combustion engineers who were trying to troubleshoot combustion problems with DLN combustion systems, but the information never played an important role in the resolution of the problem (again, unless there was NO flame present in the combustor).

So, the same flame detectors have been used for decades, with various interfaces--some of which provided an indication of intensity to the Speedtronic, many of which did not. Do the Speedtronic turbine control systems need to see intensity to determine if there's flame or not? No. The control and protection sequencing relies on a simple "ON" or "OFF" ("1" or "0") to say that diffusion flame is present. If that detection is done by looking at intensity at a very low level it's still just "ON" or "OFF".

Hope this helps!

 

Thank you CSA for your prompt and so detailed answer.
Surely that was helpful.

 

Thanks CSA for the wonderful explanation as usual. What is the difference in representing flame intensity in % and Hz?

thanks
G. Rajesh